Meta Advantage+ Is the Default Now: The 2026 UK Lead Generation Playbook

Meta Advantage+ is now the default campaign setup for sales, leads and app promotion. AI chooses audience, placements, budget split and creative variations unless you override it. Manual controls still exist but are no longer the starting point. Lead quality depends almost entirely on what you feed back from your CRM.

Every new lead campaign a UK advertiser builds in Ads Manager now starts as a Meta Advantage+ campaign. Not as an option presented alongside the manual build, as the build itself. That change happened gradually enough that plenty of teams have not registered it. They open Ads Manager, work through a setup flow that looks broadly familiar, and launch a campaign in which the audience, the placements, the budget allocation and several creative decisions have already been made by a model. The results are genuinely mixed. Some UK accounts have seen cost per lead fall by a third. Others are drowning in leads that never answer the phone, at a cost per lead that looks brilliant in Ads Manager and catastrophic in the CRM. This playbook covers what the defaults actually changed, which controls you still have and where to find them, how to tell a good lead from a cheap one, and how to structure a campaign on a UK budget under £2,000 a month.

What “default” actually changed in Ads Manager

The word default is doing a lot of work in how Meta Advantage+ has been communicated, so it is worth being precise about what moved.

The unified setup flow

Meta merged the manual campaign flow and the Advantage+ flow into a single interface. There is no longer a fork in the road where you choose automated or manual at the start. For sales, leads and app promotion objectives, the campaign you create is an Advantage+ campaign. The setup steps that used to ask you to define an audience, choose placements and set budgets per ad set now present single-step Advantage+ solutions instead. The controls were not removed. They were demoted. Most of what you used to configure at the top level now sits behind an expandable section, a toggle, or an advanced settings link that is easy to walk past.

What is switched on before you touch anything

Since February 2026, new campaigns under Sales, Leads and App Promotion launch with every Advantage+ Creative enhancement pre-selected. That covers image adjustments, text variations, music, and the newer generative features. A campaign built by clicking through the defaults will run creative that differs from what you uploaded. Advantage+ audience is on, which means your audience input becomes a suggestion rather than a boundary. Advantage+ placements is on, so your ads run across Facebook, Instagram, Messenger, Threads and the Audience Network unless you say otherwise.

What you can still override, and where

Everything meaningful remains adjustable, which is the part most coverage understates. Audience controls survive as audience suggestions plus hard exclusions. Age and location minimums still bind. Exclusions still bind absolutely, which makes them the most reliable control you have. Placement controls survive under manual placements. Turning off Audience Network alone is worth testing for most UK lead generation accounts, because it consistently produces the cheapest and worst leads. Creative enhancements can be switched off individually rather than as a block, which matters when one enhancement is helping and another is mangling your brand.

Tip: Build one campaign entirely through the defaults and one with your overrides applied, then compare them on CRM outcomes rather than on cost per lead. Most UK advertisers we work with discover that two or three specific enhancements account for nearly all the damage, and the rest are fine.

Why Meta made this change

Auction economics explain it better than any product rationale. Campaigns with wider targeting and more creative variation give Meta’s models more room to optimise, which increases the value Meta extracts from each impression. That is not a criticism, it is just an alignment problem worth naming. Meta’s model optimises towards the event you tell it to value, at the lowest cost it can find. If the event you value is a form submission, you will get form submissions, cheaply, from whoever fills forms most readily. The January 2026 auction changes reinforced this by giving Advantage+ Shopping and Lead campaigns higher priority in the auction. Running a fully manual campaign now carries a delivery cost as well as a labour cost.

Advantage+ Leads campaigns: structure, and what you give up

Meta Advantage+ Leads is now available globally as an end-to-end automation for lead generation, and it deserves a clear-eyed assessment rather than either the vendor pitch or the reflexive dismissal.

How the structure differs

A traditional lead campaign gave you a campaign, several ad sets each with its own audience and budget, and ads inside them. You controlled the split. Advantage+ Leads collapses much of that. Budget sits at campaign level and moves between audiences automatically. Audience definition becomes a signal rather than a constraint. Placements are handled for you. What you keep: the objective, the optimisation event, the creative assets you upload, your exclusions, and the form or landing page itself.

What you genuinely give up

Three things, and it is worth being honest that two of them hurt. Granular budget control is the first. You can no longer guarantee that a specific audience receives a specific amount of spend, which makes structured audience testing considerably harder. Clean audience-level reporting is the second and the more painful. When budget moves fluidly between audiences, attributing performance to an audience becomes approximate. Teams that built their optimisation routine around audience-level analysis lose their main working method. Placement-level control is the third, and it is the one you can recover most easily by switching to manual placements.

Where it performs well

Accounts with high lead volume and a clear, single conversion event tend to do well. The model has enough signal to work with and a clear target to aim at. Businesses selling something with broad appeal at a relatively low consideration threshold also do well, because the wide audience is genuinely appropriate. Local services, consumer finance enquiries, home improvement and hospitality all fit this pattern, which covers a large share of UK lead generation. Accounts where speed matters do well too. Advantage+ Leads reaches stable performance faster than a manual build with several ad sets each learning separately.

Where it performs badly

Narrow, high-value B2B audiences are the clearest failure case. A model told to find people who fill in forms will find people who fill in forms, and in a niche B2B market most of those people are not your buyers. Regulated advertisers face the creative problem described later in this guide. Financial services firms in particular should not run generative creative enhancements without review. Businesses that cannot measure what happens after the form is submitted should not run this at all. That is the single strongest predictor of a bad outcome, and it has nothing to do with Meta.

The conversion thresholds, and which one actually applies to you

There is a widely repeated claim that Meta dropped its conversion threshold to 25 per week, and UK advertisers on small budgets have taken a lot of encouragement from it. The detail matters, because the number that gets quoted is not the number that applies to lead campaigns.

The 25 per week figure

Meta reduced the threshold to 25 conversions per week in April 2026, and that change was real and significant. It applies to purchase-optimised Advantage+ Shopping campaigns. Ecommerce advertisers running purchase events benefited immediately, and for small UK online retailers it opened up a campaign type that had previously needed volume they could not reach.

The figure that applies to lead campaigns

Lead generation campaigns work to a different requirement. For stable optimisation, Advantage+ Leads generally needs somewhere around 50 lead form completions per week. That is a meaningfully higher bar, and it changes the arithmetic for a UK SME. Fifty leads a week is roughly 215 a month. At a £15 cost per lead that is around £3,200 of monthly spend before the system is optimising reliably. A business spending £800 a month on Meta lead generation is not going to hit stable optimisation, and no amount of campaign structure will change that. Knowing this before you start is the difference between a considered plan and a disappointing quarter.

What to do below the threshold

Running under the volume requirement is not pointless, but it needs different expectations and a different approach. Optimise for a higher-volume event. If lead form completions are too sparse, optimising for landing page views or a micro-conversion gives the model more signal, at the cost of precision. Consolidate ruthlessly. One campaign, one ad set, several creatives. Splitting a small budget across multiple ad sets guarantees that none of them learns. Extend your judgement window. At low volume, weekly performance is mostly noise. Monthly comparison is the minimum honest basis for a decision. Accept a longer ramp. Expect six to eight weeks before the numbers mean anything, and plan the budget so you can survive that without panic.

The honest position for small budgets

For UK businesses under about £1,500 a month, Meta lead generation can still work well, but it works through creative quality and offer strength rather than through algorithmic optimisation. The model cannot help much at that volume. What moves the needle is a genuinely compelling offer, creative that stops the scroll, and a form that does not ask for eleven fields. That is a less exciting answer than a settings change, and it is the accurate one. Our Meta Ads work with smaller UK budgets concentrates there for exactly that reason.

The 2026 timeline: what changed and when

Several changes landed across the year and they interact. Seeing them in sequence explains why accounts that were stable in January started behaving differently by summer.

Why the sequence matters

Each change looked minor in isolation and was reported that way at the time. The cumulative effect on how a lead campaign behaves has been substantial, and it arrived without any single announcement that would have prompted a review. An account built in 2025 and left alone is running under different rules from a campaign created last week, even inside the same ad account. That explains a pattern we see constantly: a business reports that Meta “stopped working” around spring, when what actually happened is that a new campaign was built through the new defaults and behaved differently from the older campaigns beside it. Auditing when each live campaign was created is a genuinely useful five-minute exercise, because creation date now predicts configuration.

The compounding effect

Higher auction priority for Advantage+ campaigns, plus enhancements on by default, plus a merged setup flow, adds up to a strong pull towards automation for anyone not actively resisting it. None of these changes was individually dramatic. Together they moved the default behaviour of the platform substantially within about six months.

Advantage+ audience: how the signal actually works

The audience change causes more confusion than any other part of the system, largely because the language Meta uses is imprecise.

Suggestion, not instruction

Under Advantage+ audience, the interests, demographics and behaviours you enter become a starting signal. Meta explores beyond them when the model expects better results elsewhere. Your input genuinely matters at the start, since it shapes early delivery while the campaign learns. It stops being a boundary once the system finds performance outside it. For advertisers who spent years building precise interest stacks, this feels like losing control. In practice the precise stacks were often producing worse results than the model does, which is uncomfortable but usually true for consumer lead generation.

What still binds absolutely

Three controls remain hard constraints rather than suggestions, and they are where your real influence sits. Exclusions bind. Anyone in an excluded audience will not be shown your ads, which makes exclusion lists the most reliable targeting tool you now have. Geographic targeting binds. If you serve Suffolk and Norfolk, setting that constrains delivery properly. Minimum age binds, which matters for age-restricted or regulated products.

Using exclusions as your primary control

Since exclusions bind and inclusions do not, the sensible response is to invert how you think about targeting. Exclude existing customers, so you stop paying to reach people who already bought. Exclude everyone currently in your sales pipeline, since they are already in conversation with you. Exclude previous form submitters after a suitable window, which prevents the duplicate lead problem described later. Exclude employees and, where you can identify them, competitors. A well-built exclusion strategy now does more for targeting precision than any interest selection, and it is the part of audience configuration most UK accounts have never touched.

Where broad targeting genuinely wins

Consumer services with wide appeal do better under broad targeting than under hand-built interest stacks, because purchase intent is poorly correlated with stated interests. Someone who needs a new boiler does not follow boiler-related pages. They become a buyer the week their boiler fails, and behavioural signals catch that far better than interest targeting ever did. For that category of business, the honest advice is to stop fighting the audience system and put the saved effort into creative and offer instead.

Website forms versus instant forms, and the lead quality difference

This is the decision that shapes lead quality more than any targeting setting, and most UK advertisers make it once and never revisit it.

How the two differ in practice

Instant forms open inside Facebook or Instagram, pre-filled with the details Meta already holds. The user taps twice and they are done. Website forms send the user to your site, where they fill in a form you built, on a page you control. The friction difference is the whole story. Instant forms convert at dramatically higher rates and produce dramatically lower cost per lead. They also produce leads who never left the app, never saw your website, and in many cases barely registered which company they enquired with.

The quality gap in numbers

Across UK lead generation accounts, the pattern is consistent enough to plan around. Instant forms typically produce two to four times the lead volume at a third to a half of the cost per lead. Contact rates tell the opposite story. Website form leads answer the phone more often, know who you are, and convert to appointments at meaningfully higher rates. Cost per lead and cost per customer routinely rank the two formats in opposite directions, which is why an account judged on Ads Manager data alone will nearly always drift towards instant forms and worse commercial outcomes.

Running both together

Meta’s own data suggests website form campaigns running alongside instant form campaigns see around 60% lower cost per lead and roughly 125% more lead volume than website forms running alone. Meta’s Advantage+ leads documentation sets out the combined approach. Treat that figure carefully. It compares a combined setup against website forms alone, so the improvement partly reflects instant forms doing what instant forms do: producing cheaper leads in greater numbers. The combination is still worth running for most accounts, provided you keep the two measurable separately and judge each on downstream outcomes rather than blended cost per lead.

Making instant forms produce better leads

Several adjustments narrow the quality gap substantially. Switch the form type from more volume to higher intent. This adds a review step before submission and cuts accidental completions sharply. Add a qualifying question that requires actual typing. One free-text field removes a surprising proportion of low-intent submissions, because it defeats the two-tap pattern that produces them. Ask for information the prospect must think about: budget range, timescale, postcode, property type. Anything that requires a decision rather than a pre-filled tap. Write the form intro as a filter rather than a welcome. Stating the price range or the minimum project size removes people who were never going to buy. Set expectations about what happens next. A line saying you will call within one working day measurably improves answer rates, because the call is expected rather than a surprise.

When to use each

Instant forms suit high-volume, low-consideration offers where speed of response is the main competitive advantage, and where your sales process can absorb a high contact-attempt workload. Website forms suit higher-value services, longer consideration cycles, and any business where sales capacity is the constraint rather than lead volume. If your sales team is already struggling to work the leads you have, more cheap leads will make your results worse, not better. That is a capacity problem and it is worth naming before adjusting any campaign setting.

Instant form questions that actually qualify

The form is the cheapest lever in the whole account and the one most UK advertisers never touch after launch. Two or three well-chosen questions change lead quality more than a month of settings work.

The principle behind a good question

A qualifying question does one of two jobs. It either gives your sales team information they would otherwise have to ring up and ask for, or it removes someone who was never going to buy. The best questions do both. Asking for a budget range tells you where the enquiry sits and simultaneously discourages people who now realise the price is not what they hoped. Questions that do neither are the ones to cut. Asking how someone heard about you is useful for reporting and useless for qualification, so it belongs in your CRM rather than on the form.

Questions that work for UK service businesses

A budget or price range question, offered as bands rather than free text. Three or four bands, with the lowest band set at or slightly below your genuine minimum. People who select the lowest band are telling you something useful. A timescale question. Immediately, within three months, within a year, or just researching. The last option is the most valuable answer on the whole form, because it lets your team deprioritise without discarding. A postcode or full address field. This filters geography properly, and for anyone doing site-based work it is information you need anyway. A free-text description of what the customer wants. One or two sentences. This is the single most effective filter available, because it defeats the two-tap pattern that produces most junk leads. A property or business type question where relevant. Detached, semi, terraced, flat, commercial. Or for B2B, company size and sector.

Questions that quietly do damage

Anything requiring information the customer does not have to hand. Exact measurements, model numbers, account references. These lose good leads alongside bad ones. Long multiple-choice lists. More than five options on a mobile form produces drop-off without producing better information. Anything that reads as intrusive before trust exists. Date of birth, income, employment status. Regulated advertisers sometimes need these, and where they do it is worth explaining why on the form itself. More than about five questions in total. Every additional question costs volume, and past five the marginal filtering value drops below the volume cost for most businesses.

Writing the form introduction

The intro is the most under-used piece of copy in Meta lead generation. Most advertisers leave it as a restatement of the ad. Use it as a filter instead. State plainly what you do, roughly what it costs, and who it suits. “We install garden rooms across Suffolk and Norfolk. Most projects run between £15,000 and £40,000 and take six to ten weeks from survey.” That intro loses a meaningful proportion of submissions and improves every one that remains. Set the expectation for what happens next in the same breath. Naming a specific response time raises answer rates, because the call arrives expected rather than as an interruption.

The completion screen

The screen after submission is another wasted asset in most accounts. It usually says thank you and nothing else. Use it to reinforce the next step, name who will call and roughly when, and give an alternative contact route for anyone who would rather ring you. Adding a calendar link here works well for businesses that can support it, since a self-booked appointment removes the contact-rate problem entirely for that lead.

Testing form changes properly

Change one thing at a time and give each change a fortnight, because form changes affect the optimisation event and therefore restart learning. Measure on contact rate rather than on submission volume. A form change that halves submissions and doubles contact rate is a clear win, and submission volume alone will tell you the opposite. Keep a record of every version. Six months in, knowing which combination of questions produced the best qualified-lead rate is genuinely valuable and impossible to reconstruct from memory.

Lead verification, and whether it fixes junk leads

Meta has expanded its lead generation tools with verification features aimed squarely at the quality complaints that have followed Advantage+ Leads since launch.

What the verification tools do

The tools work on validation and filtering: checking that contact details are plausible and real, and giving advertisers ways to screen submissions before they reach the sales team. Third-party tools cover similar ground, with email validation and phone verification available from several established providers. These help with one specific problem: fake or mistyped contact details. A lead with a real email address and a real phone number is a genuine improvement over one with neither.

What they do not fix

Verification confirms that a person exists and that their details are correct. It says nothing about whether they want to buy anything. Coverage of the verification rollout, including the CRM expansion that accompanied it, is summarised well by Social Media Today’s write-up of the lead gen updates, which is worth reading alongside Meta’s own materials for a less promotional framing. The dominant junk lead problem in UK accounts is not fake data. It is real people with real phone numbers who tapped twice out of idle curiosity and have no memory of doing it. No verification tool addresses that, because there is nothing invalid to detect. Treat verification as hygiene rather than as a quality strategy. It removes an irritation. It does not change what the algorithm is optimising towards.

The actual fix

Lead quality is determined by the optimisation event, and changing the optimisation event is the only intervention that reliably works. If Meta is optimising for form submissions, it will find the people most likely to submit forms. If it is optimising for an event that only fires when a lead reaches a qualified stage in your CRM, it will find people more likely to reach that stage. That is the whole mechanism, and everything else is a refinement around it. The configuration is covered later in this guide, and it is the single highest-value change available to most UK Meta advertisers.

A warning about optimising for cost per lead

Judging a Meta lead campaign on cost per lead creates a direct incentive for the system to find you the cheapest possible leads. It is very good at this. Practitioners have put it more bluntly: if you measure by cost per lead, the AI will happily deliver £2 leads that are entirely worthless. The uncomfortable version of this is that a campaign getting worse commercially will often look like it is improving in Ads Manager, and it can continue looking that way for months if nobody checks the CRM.

Advantage+ Creative in 2026, and what the AI is doing to your ads

The creative side has moved faster than the targeting side, and it is where the defaults do the most unexpected work.

What is now generated automatically

Advantage+ Creative applies a set of enhancements that were, until recently, individually optional and are now pre-selected on new campaigns. Background generation extends or replaces the background of a static image. Video generation produces short video from still assets. Headline variation rewrites your copy into alternatives. Text length optimisation trims or expands copy per placement. AI-generated music can be added to video, matched to the tone of the creative. AI dubbing adapts spoken audio into other languages, part of the set of AI advertising features Meta has rolled out through 2026. Meta announced Muse Image in July 2026, built by its Superintelligence Labs team, with Advantage+ Creative integration following for larger accounts. Meta’s own Advantage+ Creative page documents the current feature set.

Which enhancements are usually safe

Text length optimisation is generally harmless and often helpful, since it adapts copy to placement constraints you would otherwise handle manually. Brightness and visual adjustments rarely cause problems for consumer brands, though they can shift a carefully specified brand colour. Placement-specific cropping is usually a net positive, because manually producing every aspect ratio is exactly the kind of work worth automating.

Which enhancements need watching

Background generation is the most common source of complaints. Generated backgrounds can misrepresent a product, invent a setting that does not exist, or produce visual artefacts that look cheap. Headline variation rewrites your copy. For a regulated advertiser, that means claims nobody approved. For everyone else, it means brand voice drifting away from whatever you agreed in your tone of voice guidelines. Video from static images is covered in its own section below, because it is the enhancement with the widest quality range.

Approval and control

Advertisers retain approval control over generated content before it goes live, which is the reassurance worth knowing about and acting on. The practical problem is that approval only happens if someone looks. Enhancements are on by default, generated variants appear as the campaign runs, and a busy team can go weeks without reviewing what is actually being served. Build the review into a routine rather than relying on noticing. Fifteen minutes a fortnight looking at live creative variants catches nearly everything worth catching.

Video from static images: where it works and where it looks cheap

This enhancement generates short video from a still photograph, usually by adding motion, parallax or a slow zoom. It is the most variable of the creative features, and the results range from genuinely useful to actively damaging.

Where it works

Product photography on plain backgrounds handles motion well. A subtle zoom or rotation adds movement to the feed without distorting anything. Lifestyle photography with depth works too, because parallax needs foreground and background separation to look natural. A photograph with a clear subject and a receding background will usually animate convincingly. Accounts with limited creative production budgets get the most value. Turning a library of twelve product photographs into twelve short videos is real work saved, and video placements reach inventory that static images do not.

Where it looks cheap

Photographs of people are the main problem. Generated motion applied to faces produces the uncanny effect that makes a viewer distrust the advert without knowing why. Detailed or text-heavy images degrade badly. Anything with fine detail, small text or precise geometry tends to warp during motion generation. Premium positioning suffers most. A brand selling on craftsmanship or quality is undermined by creative that looks automatically generated, and UK consumers have become noticeably quicker at spotting it.

Testing it properly

Run it as a genuine test rather than leaving it on by default. Duplicate an ad set, enable video generation on one side, and compare on downstream outcomes rather than on click-through rate. Watch the generated videos yourself on a phone, in the feed, at normal scrolling speed. Reviewing creative on a desktop monitor at full size misses almost everything a real viewer notices. Judge on lead quality as well as volume. Generated video often lifts engagement while producing weaker leads, because motion attracts attention from people who were not otherwise interested. If you would not have signed off the creative when it was produced by a person, do not run it because it was produced by a model. The standard is the same either way.

Feeding real outcomes back: CRM conversion events instead of form fills

Everything in this guide leads here. This is the change that separates accounts where Meta Advantage+ works from accounts where it quietly wastes money.

The problem in one sentence

Meta optimises towards the event you tell it to value, and most UK advertisers are telling it to value a form submission, which is not a commercial outcome.

What to send back instead

Send events that represent genuine progress in your sales process, not activity. A qualified lead event, firing when someone reaches a lead status that means a human has spoken to them and confirmed they are a real prospect. An appointment or meeting booked event, which for most UK service businesses is the first point where the lead has demonstrated real intent. A closed-won event with the deal value attached, which is the strongest signal available and the one that lets Meta optimise towards revenue rather than volume.

How it works technically

The Conversions API handles CRM uploads, and it is a separate integration from a standard website Conversions API setup because the parameters and data source differ. Meta’s developer documentation on conversion leads integration covers the implementation. Paired with the Conversion Leads performance goal, this shifts optimisation from form completions towards leads that actually progress. Most UK businesses will do this through their CRM’s native integration rather than building against the API directly. We have written up the HubSpot approach in detail in optimising Meta campaigns for HubSpot CRM conversion events, and the same principles carry to other CRMs.

The attribution window constraint

There is a timing limitation that catches people out. Events uploaded more than seven days after the ad click may not attribute correctly. For a business with a two-week sales cycle, closed-won is therefore useless as an optimisation event, because the deal closes long after the attribution window has passed. The workaround is to send the earliest event that still indicates quality. For most UK service businesses that means a qualified lead status or a booked appointment, both of which typically occur within a few days. Choose the highest-quality event that reliably happens inside seven days. That single decision determines how well the whole system performs.

Building better audiences from the same data

Once qualified events flow back, you can build custom and lookalike audiences from leads who actually converted rather than from everyone who submitted a form. The difference in seed quality is substantial. A lookalike built from 500 customers behaves very differently from one built from 5,000 form fills of unknown value. This is the compounding benefit, and it is why the CRM integration is worth doing even for accounts too small to hit the optimisation thresholds discussed earlier.

Reading Meta’s uplift figures sceptically, and testing them yourself

Meta publishes performance claims for Advantage+ features, and more than eight million advertisers were using its AI creative tools as of July 2026. Adoption is not evidence of performance.

How to read platform-published figures

Vendor performance figures are usually accurate and usually not applicable to you, for three reasons worth understanding. The comparison group is self-selected. Advertisers who enable a feature and keep it on are those for whom it worked, which builds the result into the measurement. The metric chosen tends to favour the feature. Cost per lead improvements are easy to demonstrate and say nothing about lead quality. Aggregate figures hide variance. A median improvement of 20% is compatible with half the accounts getting worse.

Designing a test you can trust

Use a holdout rather than a before-and-after. Comparing this month to last month attributes seasonality, market shifts and your own other changes to the feature you happen to be testing. Split at the campaign level with matched budgets, run both simultaneously, and leave them alone for the full window. Define the success metric before you start, and make it a CRM metric. Cost per qualified lead or cost per appointment, not cost per lead. Set the window by volume, not by calendar. You need enough conversions on both sides for the difference to mean something, which at low volume can take six weeks or more.

The metrics worth tracking

Four numbers tell you nearly everything about a UK lead generation account on Meta. Cost per qualified lead, meaning leads a human has confirmed as genuine prospects. This is your primary number. Contact rate, the proportion of leads you can actually reach. This is the fastest early warning of a quality problem, and it moves within days. Lead-to-appointment rate, which tells you whether the leads reaching you are the right people. Cost per appointment or per closed deal, which is the number your finance director cares about and the only one that reflects commercial reality. Cost per lead belongs on the list only as a diagnostic. When it falls sharply while contact rate falls with it, you are looking at a quality problem rather than an efficiency gain. Improving that picture is usually a conversion rate optimisation problem as much as a media buying one.

Creative that works for UK lead generation in 2026

With Meta Advantage+ handling targeting and placements, creative is where the remaining advantage sits. It is also where most UK accounts are weakest.

Why creative carries more weight than it used to

When you could hand-build audiences, a mediocre advert shown to precisely the right people performed acceptably. That lever has mostly gone. Now the model decides who sees the advert, and it makes that decision largely from how people respond to the creative itself. Creative has become the targeting mechanism rather than the message delivered after targeting. The practical implication: your creative now selects your audience. An advert that appeals to bargain hunters will be shown to bargain hunters, and no setting will correct for that.

The four concepts worth having live

Most UK lead generation accounts do well with four distinct concepts running together, each doing a different job. A problem-led concept that names the specific frustration the customer has, in their own words rather than industry language. A proof concept using a real result, a real customer or a real before-and-after. UK audiences are markedly sceptical of unsubstantiated claims, and specificity beats superlatives. A filter concept that states price range, minimum project size or qualifying criteria. This one deliberately reduces volume and improves everything downstream. An objection concept that addresses whatever stops people buying: disruption, timescale, hidden costs, or whether you cover their area.

What consistently underperforms

Stock photography of people who obviously are not your customers. UK audiences spot it instantly and it damages credibility more than having no image. Copy that leads with the company rather than the customer. Nobody scrolling is interested in how long you have been established. Vague offers. “Get in touch for a free quote” attracts people who collect quotes, which is a genuinely different population from people who buy. Over-produced video for local services. Slightly rough authentic footage routinely outperforms polished production for trades and home improvement, because it reads as real.

Refresh rate

Creative fatigue is faster than most teams expect, particularly at small budgets where the reachable audience is limited. Plan on refreshing concepts monthly and individual assets fortnightly. That sounds demanding until you compare it against the cost of a campaign quietly declining for six weeks. Keep winners running while introducing challengers, rather than replacing everything at once. Wholesale creative changes restart learning and destroy the comparison.

Writing for the filter

The most useful creative habit for a capacity-constrained business is writing adverts that repel the wrong people as deliberately as they attract the right ones. Naming a price range does this better than anything else. So does specifying the area you cover, the minimum job size, or the type of customer you work with. It feels counterintuitive to advertise in a way that reduces response. For a business where sales time is the constraint rather than lead volume, it is the highest-return change available.

A budget and structure framework for UK campaigns under £2,000 a month

Most published Meta advice assumes budgets that UK SMEs do not have. This section assumes you have between £500 and £2,000 a month and need it to work.

The structure that suits small budgets

Consolidate. One campaign, one ad set, four to six creatives is the right starting structure for almost every account at this level. Every additional ad set divides your conversion volume, and conversion volume is the scarce resource. Three ad sets on a £1,000 budget produce three under-optimised ad sets rather than one working one. Let the campaign run for at least three weeks before judging anything. Restarting the learning phase by making changes is the most common self-inflicted problem at small budgets.

Budget allocation by tier

At £500 to £800 a month, run a single campaign with instant forms, optimise for lead form completion, and concentrate effort on creative and offer. Algorithmic optimisation is not available to you at this volume, so the only real lever you have is the creative. At £800 to £1,500, add a website form campaign alongside the instant form campaign, split roughly 70/30 in favour of instant forms. Begin sending qualified lead events back from your CRM even though volume is low, because the audience-building benefit starts immediately. At £1,500 to £2,000, move the optimisation event to qualified lead if your CRM integration is working and you are generating at least 40 leads a month. Expect cost per lead to rise and cost per qualified lead to fall.

What to spend on creative

The uncomfortable ratio: at small budgets, creative quality matters more than everything else in this guide combined. Four to six genuinely different creative concepts, refreshed monthly, will outperform any settings optimisation available to you. Different concepts, not four crops of the same photograph. Include at least one piece of creative that names the price or the qualifying criteria. Filtering at the creative stage costs you volume and improves everything downstream, which is the correct trade at a small budget where sales capacity is limited.

The offer matters more than the targeting

Meta’s model will find people who respond to your offer. If the offer is weak, it will find people who respond to weak offers, which is a small and unpromising group. Strong offers for UK lead generation tend to share features: a specific outcome, a clear timescale, a reason to act now, and a low-commitment first step that is still meaningful. “Free quote” is not an offer. “Fixed-price survey within five working days, deducted from your project cost if you proceed” is an offer, and it will outperform any audience adjustment.

A realistic timeline

Weeks one to three are learning. Cost per lead will be high and variable, and the correct response is to leave it alone. Weeks four to six give you a first honest read. This is when you judge creative, not settings. Weeks seven to twelve are where you make structural decisions with data behind them. Anything before week four is guesswork dressed as analysis.

When to override the defaults

The defaults suit a particular kind of advertiser. Several common UK profiles should change them, and knowing which one you are saves a lot of wasted spend.

Regulated advertisers

Financial services firms, mortgage and insurance brokers, solicitors and healthcare providers all face the same problem: generative creative produces claims nobody approved. Switch off headline variation and text generation entirely. Keep placement optimisation and cropping, which carry no claim risk. Review every live creative variant weekly for the first month. The FCA, SRA and ASA do not distinguish between copy your team wrote and copy a model wrote on your behalf. Consider website forms over instant forms, since they give you a page where required disclosures can sit properly rather than being compressed into a form intro.

High-value B2B advertisers

Broad audiences find people who fill in forms, and in B2B that is rarely the decision maker. Use exclusions aggressively, since exclusions bind absolutely where audience suggestions do not. Exclude existing customers, competitors, current pipeline and any job function irrelevant to the purchase. Optimise for a qualified event as early as your CRM allows, and accept that cost per lead will look poor. Set expectations before launch. A B2B account should expect a cost per lead several times higher than a consumer account, and judging it on consumer benchmarks produces bad decisions.

Premium and luxury brands

Generative creative enhancements are the main risk, particularly background generation and video from static images. Switch off anything that alters imagery. Keep the enhancements that adapt format rather than content. Audience Network placement is worth excluding, since the inventory quality is inconsistent with premium positioning.

Accounts with limited sales capacity

More leads is not automatically better, and for a small team it is often worse. Deliberately raise friction. Higher intent form types, qualifying questions and price-anchored creative all reduce volume and improve what arrives. Optimise for appointments rather than leads if your CRM supports it, which aligns the system with your actual constraint.

Businesses with strong first-party data

Accounts with a substantial customer list have an advantage worth using rather than leaving to broad targeting. Upload customer lists, build lookalikes from your best customers rather than all customers, and use value-based lookalikes where you have deal values. This is the one area where a small UK advertiser can genuinely outperform a bigger competitor, because the quality of your first-party data is not a function of your budget.

The structure and settings checklist

A practical list to work through before launching, or to audit an existing account against.

Campaign level

1. Objective set to Leads, with the optimisation event chosen deliberately rather than accepted.
2. Budget at campaign level, sized so the campaign can plausibly reach its optimisation threshold.
3. Conversions API connected and passing at least one qualified event back from your CRM.
4. Exclusions applied: existing customers, current pipeline, employees, competitors.

Ad set level

1. Placements reviewed, with Audience Network tested off for lead generation.
2. Age and location minimums set to your actual serviceable market, not aspirationally wide.
3. Attribution setting checked and consistent with how you report internally.
4. Schedule set if your sales team cannot respond outside working hours, since lead response time is the strongest predictor of contact rate.

Creative level

1. Every Advantage+ Creative enhancement reviewed individually and switched off where inappropriate.
2. Four to six genuinely distinct concepts live, not variations of one.
3. At least one creative that filters by naming price, scope or qualifying criteria.
4. A review date in the calendar for checking generated variants.

Form level

1. Form type set to higher intent rather than more volume for most UK service businesses.
2. At least one question requiring free text or a real decision.
3. Intro copy written as a filter, stating what you do and roughly what it costs.
4. Expectation set for what happens next and when.
5. Lead delivery tested end to end into your CRM, with response time measured from submission rather than from when someone noticed.

A worked example: a £1,800 a month UK home improvement business

Here is the arithmetic on a profile we see repeatedly, because abstract advice about lead quality becomes much clearer with numbers attached.

The starting position

A regional installer covering East Anglia, average job value around £6,500, two people handling enquiries alongside their other work. The account runs one Advantage+ Leads campaign with instant forms, built through the defaults, optimising for lead form completion. Monthly spend £1,800. Ads Manager shows 240 leads at £7.50 each. On paper that is an excellent result and the business was pleased with it for two months.

What the CRM showed

Of those 240 leads, the team managed to reach 68. Contact rate 28%. Of the 68 contacted, 19 agreed to a survey. Of those, 6 became customers. Cost per customer: £300, against a £6,500 job value. Still profitable, and the business was right to be reasonably happy. The problem was not profitability. It was that two people spent most of their week chasing 172 leads who never answered, which capped how much the business could actually grow.

The changes made

Four changes, applied over six weeks rather than at once. Form type switched from more volume to higher intent, and a free-text question added asking what work the customer wanted doing. Creative rewritten so one of the four concepts stated a typical project range of £4,000 to £12,000. Qualified lead events connected from the CRM, firing when a team member marked a lead as a genuine prospect after a conversation. Audience Network switched off after a test showed it producing 22% of leads at a 9% contact rate.

The result after eight weeks

Leads fell from 240 to 96. Cost per lead rose from £7.50 to £18.75, which looked like a significant deterioration in Ads Manager. Contact rate rose from 28% to 61%. The team reached 59 leads from 96, almost the same absolute number as before from 40% of the volume. That last point is the one worth sitting with: the business lost 60% of its leads and lost almost none of the conversations it actually wanted. Surveys booked rose from 19 to 26. Customers rose from 6 to 9. Cost per customer fell from £300 to £200, on the same budget, while the team’s workload dropped by more than half. Every headline metric in Ads Manager got worse and the business got materially better.

What made the difference

Sequencing mattered. The qualified lead event was connected before the form changes, so Meta had a better target to optimise towards as the volume dropped. Patience mattered more. Cost per lead more than doubled in week two and the honest instinct was to revert. Holding the change through the learning phase was the hardest part and the most important. The general lesson: when you improve lead quality, your platform metrics get worse first and your business metrics improve later. Anyone judging the change at week three would have reversed it.

Troubleshooting the problems that show up most

These are the issues UK accounts hit most often, with the diagnosis that usually applies.

Cost per lead is brilliant and nobody answers the phone

The most common problem in Meta lead generation, and almost always caused by optimising for form completions with a low-friction instant form. Check contact rate first. Below about 35% you have a quality problem regardless of what cost per lead says. Fix by raising form friction, adding qualifying questions, and moving the optimisation event to something further down your funnel. Expect volume to fall by half or more, and expect that to be the correct outcome.

Lead volume collapsed after a change

Usually the learning phase rather than a broken campaign. Any change to the optimisation event, audience or budget restarts learning. Wait a full two weeks before reacting. Changing something else during the learning phase compounds the problem and restarts the clock again. If volume has not recovered after three weeks, look at whether your optimisation event fires often enough to sustain delivery. An event happening twice a week cannot drive a campaign.

Generated creative that misrepresents the business

Switch off the specific enhancement rather than all of them. Background generation and headline variation cause most complaints; format adaptation rarely does. Review what has already run. Generated variants may have been serving for weeks, and for regulated advertisers that is worth documenting properly rather than quietly fixing.

Leads arriving hours late

A CRM integration problem rather than a Meta one, and it damages results more than most targeting decisions. Contact rate falls sharply with response time. A lead called within five minutes is dramatically more likely to answer than one called the next morning. Test the full path from form submission to a phone ringing, and measure it. Teams routinely discover a delay nobody knew about.

The same leads appearing repeatedly

Deduplicate in the CRM and add submitters to an exclusion audience. Without exclusions, someone who enquired last month remains eligible to see your ads and enquire again, and Meta counts that as a fresh conversion.

Performance good on Facebook, poor on Instagram, or vice versa

Break the campaign out by placement in reporting before restructuring anything. Resist splitting into separate campaigns per platform at small budgets, since that divides conversion volume. Use placement exclusions within one campaign instead.

Working the leads: speed, sequence and what to say

The best campaign in the world feeds a process, and for most UK businesses the process is where the money leaks. This section sits outside Ads Manager entirely and it will improve your results more than anything inside it.

Speed is the whole game

Contact rate falls sharply with elapsed time, and the drop is steepest in the first hour rather than across the first day. A lead called within five minutes is dramatically more likely to answer than the same lead called three hours later. By the next morning you are ringing someone who has forgotten enquiring and may already have spoken to a competitor. That matters more with instant forms than with website forms, because the enquiry took two taps and carried almost no commitment. The window in which the person still remembers doing it is short. If you can only fix one thing after reading this guide, fix the time between form submission and the first attempted contact. It costs nothing in media spend.

Building a contact sequence that works

One call is not a process. Most businesses give up far too early and write the lead off as junk when it was simply missed. A workable sequence for a UK service business: call within five minutes, text immediately if unanswered, call again two hours later, call the following morning, then a final call two days on. Five touches across three days. Vary the time of day across attempts. Someone unreachable at 10am on a Tuesday may answer easily at 6pm. Send something in writing after the first unanswered call. A short text naming your company and referencing what they enquired about converts a fair number of unanswered calls into returned ones, because it turns an unknown number into a recognised one.

What to say on the first call

Open by naming the enquiry rather than the company. “You asked about a garden room on Facebook about ten minutes ago” gets a far better reception than “I’m calling from Company X.” That framing matters for instant form leads specifically, since the person may genuinely not remember and needs the reminder before they can engage. Qualify in the first two minutes using whatever the form told you, then either book the next step or close the lead out politely. The most expensive habit in lead handling is keeping unqualified enquiries alive because nobody wants to write them off. Ask for the next step explicitly. A survey, a call with a colleague, a site visit. Leads that end without a scheduled next action rarely progress.

Recording outcomes so the campaign can learn

Everything in the CRM section of this guide depends on someone marking the outcome accurately and promptly. Keep the statuses few and unambiguous. Reached and qualified, reached and not qualified, not reached, and booked. Four options that a busy person can apply correctly in one click. Set the expectation that outcomes get recorded the same day. A qualified event that fires four days late may fall outside the seven day attribution window, at which point the campaign learns nothing from it. This is the point where lead generation stops being a marketing activity and becomes an operational one. Businesses that accept that tend to get much more from the same budget.

Capacity as a real constraint

There is a volume of leads above which your results get worse, and most small teams find it sooner than they expect. Two people handling enquiries alongside other duties can work perhaps sixty to eighty leads a month properly. Beyond that, response times slip, follow-up sequences get abandoned, and contact rate falls regardless of lead quality. Work out your honest capacity number before increasing spend. If the number is eighty and you are generating two hundred, more budget will not help and better filtering will. The uncomfortable version: a business at capacity should spend on sales resource before it spends on more media, and no amount of campaign optimisation changes that arithmetic.

Feeding the process back into the campaign

Everything your sales team learns should end up shaping the adverts, not just the CRM. The objections they hear most often are your next creative concept. The questions they answer repeatedly belong on the form or in the ad copy. Run a fifteen minute conversation with whoever handles enquiries every month. It is the cheapest research available and it consistently produces better creative direction than any amount of platform data.

How GLO approaches Meta lead generation

We do not start with Ads Manager, because the settings are the last decision rather than the first.

The Growth Strategy comes first

Before building a campaign we work through the business itself: brand archetypes, tone of voice, commercial objectives, the challenges in the way, and which audiences are genuinely most likely to buy. For each audience we build vertical profiles covering the job titles or household profiles that make the decision, the pain points behind it, the objections that stall it, and the lead magnets that actually attract enquiries. That work produces the offer, and the offer determines results far more than any Advantage+ setting. A campaign built on a weak offer cannot be rescued by configuration.

Then the CRM architecture

Lead status and lifecycle stages get built before any spend goes live, so there is somewhere for quality data to live from day one. Our standard structure runs lead status from New and Untouched through Attempted to Contact, Active Conversation, Meeting Booked and onwards to Quote Sent and Invoice Paid. Lifecycle stages track Suspect through to Client. Those stages are what feed back into Meta as qualified events. Without them there is nothing to optimise towards except form fills, which is where most underperforming accounts are stuck.

Then the campaign

Only then do the defaults get reviewed, the enhancements switched on or off, and the forms built. By that point most decisions have already been made by the strategy rather than being guessed in the interface, which is the difference between a campaign that improves over time and one that produces cheap leads nobody wants. This approach shapes how we run lead generation across paid channels, and Meta is where it makes the largest measurable difference.

A 90-day plan for moving to quality-based optimisation

This is the sequence we use when taking over a Meta Advantage+ account that has been running on form-fill optimisation, which describes most UK Meta lead generation accounts.

Days 1 to 14: measure before changing

Resist changing anything. You need a baseline and you cannot reconstruct one later. Record contact rate, lead-to-appointment rate and cost per customer for the past 90 days. If your CRM cannot produce those numbers, fixing that is the first project and everything else waits. Audit the creative that is actually serving, including generated variants. Note anything that misrepresents the business. Export current form questions, placements, exclusions and optimisation events, so you know precisely what you changed later.

Days 15 to 30: connect the CRM

Get qualified lead events flowing back to Meta through the Conversions API before touching campaign settings. Choose the event carefully. The highest-quality outcome that reliably occurs within seven days of the click, which for most UK service businesses is a qualified lead status or a booked appointment. Verify events arrive and match. Discrepancies between CRM counts and Meta’s received events are common and worth resolving before you rely on the data. Leave the campaign optimising on form fills during this period. You are building the data foundation, not changing the campaign yet.

Days 31 to 45: fix the form and the creative

Now raise friction, one change at a time. Switch form type to higher intent. Add a qualifying free-text question. Rewrite the intro as a filter. Refresh creative so at least one concept states price or scope. Switch off generative enhancements that alter imagery or claims. Expect cost per lead to rise and volume to fall. This is the change working, not failing. Track contact rate weekly, since it should start improving within days even while cost per lead looks worse.

Days 46 to 60: change the optimisation event

Switch the campaign to optimise for your qualified event, assuming it now fires at a workable frequency. Expect a full learning phase. Two weeks of unstable delivery is normal and reverting during it wastes the previous six weeks of work. If your qualified event fires fewer than about 15 times a week, stay on form fills and rely on the form and creative filtering instead. Optimising towards an event that rarely happens starves delivery.

Days 61 to 90: rebuild audiences and consolidate

Build lookalikes from qualified leads and customers rather than from all form submitters. Add all previous submitters to an exclusion audience so you stop paying to re-acquire the same people. Consolidate any campaign fragmentation that crept in, since you now have better signal and want it concentrated. Review against the day one baseline on business metrics only. Cost per lead is expected to be worse. Cost per customer is the test.

What good looks like at day 90

Contact rate above 50% for instant forms, or above 65% for website forms. Cost per qualified lead stable or improving, even where cost per lead has risen substantially. Cost per customer improved against baseline, which is the only number that settles the argument. Sales team workload down, or at least redirected from chasing unreachable leads towards conversations with real prospects.

Measurement: the reporting that keeps this honest

Good decisions need a reporting setup that shows commercial outcomes rather than platform activity.

Build the report around the funnel, not the platform

A useful weekly report has five rows: leads, contacted, qualified, appointments, customers. Alongside each, the cost per unit at that stage. Reading down that column shows where the money is genuinely going. Reading only the top row is how accounts end up optimising for the wrong thing for months. Add the previous period alongside so movement is visible without anyone having to remember last month’s numbers.

Separate the formats

Instant form leads and website form leads should never be blended into a single cost per lead. They behave differently enough that the blended number is meaningless. Report them as separate lines throughout the funnel. The comparison is usually the most instructive thing in the whole report.

Watch contact rate as the leading indicator

Contact rate moves within days of a quality change, while cost per customer takes a full sales cycle to respond. That makes it the early warning system. A falling contact rate means quality is degrading, whatever the cost per lead is doing. Put contact rate on the front page of the report, next to cost per lead, and the two of them together will tell you almost everything.

Review generated creative on a schedule

Add a standing fortnightly slot to look at live creative variants on a phone. Record what you saw and whether you approved it. For regulated advertisers this is a compliance record as much as a quality check.

Attribution honesty

Meta’s reported conversions and your CRM’s lead count will not match, and chasing perfect reconciliation wastes time. Pick one source of truth for commercial decisions, which should be your CRM, and use Meta’s numbers for in-platform optimisation only. Document the gap once, explain it to whoever reads the report, and stop relitigating it monthly.

Ten mistakes UK advertisers are making with Advantage+

Drawn from the UK accounts we have audited since the Meta Advantage+ defaults changed.

1. Judging campaigns on cost per lead. The single most expensive mistake available, because it actively rewards the algorithm for finding worse leads.
2. Leaving every creative enhancement on without looking. Defaults are chosen for Meta’s average advertiser, who is not you.
3. Splitting a small budget across several ad sets. Guarantees none of them reaches optimisation volume.
4. Expecting the 25-conversion threshold to apply to lead campaigns. It applies to purchase-optimised Shopping campaigns; lead campaigns need considerably more.
5. Reverting changes during the learning phase. Quality improvements always look like failures in week two.
6. Never connecting the CRM. Leaves the system optimising towards form fills forever.
7. Using “free quote” as an offer. Attracts people collecting quotes, not people buying.
8. Ignoring response time. A brilliant campaign feeding a slow sales process produces poor results that get blamed on the campaign.
9. Running Audience Network without testing it. Frequently a large share of volume at a fraction of the quality.
10. Not excluding previous submitters. Paying repeatedly to acquire the same uninterested people.

The through-line is that most Advantage+ problems are not Meta problems. They are measurement problems that Meta’s automation makes more expensive.

Where this is heading

Two directions are already visible, and both favour advertisers who prepare rather than react. Neither depends on guessing what Meta ships next, which is why they are worth planning around. Control keeps moving from the advertiser to the model. Manual campaign construction is becoming the exception, and the remaining controls are increasingly about constraints and signals rather than direct instruction. Creative production keeps moving into the platform. With eight million advertisers using Meta’s AI creative tools, generated assets are becoming the norm, which raises the value of creative that is distinctly yours. The advertisers who do well from here are the ones who can tell Meta precisely what a good outcome looks like. That capability lives in your CRM and your sales process, not in Ads Manager, and it is the part no platform change can take away from you.

Frequently asked questions

Can I still run manual Meta campaigns in 2026?

Yes, though the starting point has changed. Meta merged the manual and Advantage+ flows into one interface for sales, leads and app promotion objectives, so new campaigns begin as Advantage+ campaigns with automated settings pre-selected. The manual controls still exist behind advanced settings, expandable sections and toggles, and you can override audience, placements and each creative enhancement individually. What you cannot do is start from a blank manual build the way you could previously. There is also a delivery consideration: since January 2026, Advantage+ Shopping and Lead campaigns receive higher priority in the auction, so fully manual campaigns can face a cost disadvantage as well as taking more work to manage.

Does the 25 conversions per week rule apply to my lead campaign?

Almost certainly not. The 25-conversion threshold Meta introduced in April 2026 applies to purchase-optimised Advantage+ Shopping campaigns, and it was a genuine improvement for small ecommerce advertisers. Lead generation campaigns work to a different requirement, generally needing around 50 lead form completions per week for stable optimisation. That is roughly 215 leads a month, which at typical UK costs means a budget well above what most SMEs allocate to Meta. Below that volume, consolidate into a single ad set, judge performance monthly rather than weekly, and concentrate your effort on creative and offer rather than on settings.

Why are my Meta leads so poor quality?

Usually because the campaign is optimising for form submissions, and Meta is very good at finding people who submit forms. Those are not the same people as those who buy. Instant forms make it worse by removing nearly all friction, so a two-tap submission from someone idly scrolling counts identically to a considered enquiry. The fix is to change what you are optimising towards. Send qualified lead or appointment events back from your CRM through the Conversions API, switch the form type to higher intent, add a question requiring free text, and state price or scope in at least one creative. Expect volume to fall sharply and cost per lead to rise. Contact rate and cost per customer are the numbers that should improve.

Should I use instant forms or website forms?

Both, if your budget allows, kept separately measurable. Instant forms produce far more leads at far lower cost, since they open in-app and pre-fill the user’s details. Website forms produce fewer, more expensive leads who reached your site, saw your business properly and convert at higher rates. Meta’s data suggests running both together lowers cost per lead by around 60% and lifts volume around 125% against website forms alone, though that partly reflects instant forms doing what they do. Choose based on your constraint: if you need more volume and can handle the contact workload, instant forms; if sales capacity is your limit, website forms with qualifying questions.

Do Meta’s lead verification tools solve the junk lead problem?

Partially, and less than the name suggests. Verification checks that contact details are real and plausible, which removes mistyped and fake submissions. That is a genuine improvement and worth switching on. It does not address the dominant problem in UK accounts, which is real people with real contact details who submitted a form without meaningful intent and do not remember doing so. Nothing about that data is invalid, so nothing flags it. Verification is hygiene rather than a quality strategy, and the only intervention that reliably changes lead quality is changing the event the campaign optimises towards.

Where this leaves UK advertisers

Meta Advantage+ being the default is not something to fight, and the accounts doing best have stopped trying. What matters now:

• Judge campaigns on cost per qualified lead and cost per customer, never on cost per lead alone
• Connect your CRM and send back the best-quality event that reliably fires within seven days
• Review every creative enhancement individually rather than accepting the pre-selected set
• Raise form friction deliberately if your constraint is sales capacity rather than lead volume
• Consolidate small budgets into one campaign and one ad set, and leave them alone long enough to learn
• Expect platform metrics to get worse when lead quality improves, and hold your nerve through it

The automation is genuinely capable, and it will pursue whatever target you set with more persistence than any human media buyer. Setting the wrong target has therefore become considerably more expensive than it used to be. There is a version of this that sounds like a complaint about automation, and it is not. The automation is a genuine improvement for advertisers who have done the measurement work, and it removes a great deal of manual effort that never added much value. What it removes is the margin for vagueness. An advertiser who could previously get away with fuzzy targets and manual correction now finds those fuzzy targets pursued relentlessly and expensively. Everything worth doing here comes back to the same question: can you tell the difference between a lead and a customer, and can you communicate that difference to the platform? Businesses that can are getting better results than they did before the defaults changed. Businesses that cannot are getting cheaper leads and less revenue. If you want help working out which of those describes your account, book a Growth Strategy session and we will go through the numbers with you.