The learning-phase floor and why it changes everything
The single most useful framework for budget decisions is not a percentage of revenue or a sector benchmark. It is the platform's minimum viable data requirement.
Both Meta and Google Ads use machine learning to find the right customers for your business. That process requires enough conversion data to work from. Meta needs approximately 50 optimisation events per ad set per week. Google's Smart Bidding strategies need roughly 30 to 50 conversions per month per campaign to function.
Below these thresholds, the algorithm cannot optimise. It makes random or regressive decisions. You are paying for data the system cannot yet use. Performance will be inconsistent regardless of how good your creative or targeting is.
For typical UK conversion rates, these thresholds translate to approximate monthly minimums:
per ad set per month
per campaign per month
minimum to exit learning
These are floors, not targets. They represent the point below which most UK businesses will not generate enough conversions for the platform to learn. Spending less is not always wrong, but you should expect inconsistency and understand why.
Budget ranges by channel for UK businesses
These ranges reflect what UK businesses typically need to spend to get consistent, optimising results from each channel. "Minimum viable" is the floor at which the platform can begin learning; "effective range" is where most accounts operate once established.
| Channel | Minimum viable (UK) | Typical effective range | Key constraint |
|---|---|---|---|
| Meta Ads | £700 per ad set/mo | £1,500 to £8,000/mo | Learning phase requires 50 events/week |
| Google Search | £600 per campaign/mo | £1,000 to £10,000/mo | CPC varies sharply by sector |
| Google Shopping | £500/mo | £800 to £5,000/mo | Requires quality product data feed |
| Google Performance Max | £1,000/mo | £2,000 to £15,000/mo | Needs existing conversion history to work |
| LinkedIn Ads | £1,500/mo | £3,000 to £10,000/mo | CPCs 5 to 10x higher than Meta; B2B only |
| TikTok Ads | £1,000/mo | £1,500 to £6,000/mo | Requires fresh creative very frequently |
| YouTube / Connected TV | £2,000/mo | £3,000 to £15,000/mo | Video assets required; brand awareness play |
Work backwards from unit economics, not forwards from revenue
The percentage-of-revenue rule ("spend 10% of turnover on marketing") is widely repeated and routinely wrong for growing businesses. It describes what you are spending today, not what you should spend to grow.
The unit-economics approach works differently:
- What is your target cost per acquisition (CPA) or cost per lead (CPL)?
- How many new customers or leads do you need per month?
- What is your realistic conversion rate from ad click to sale or enquiry?
Multiply the required number of customers by your target CPA and you have a minimum ad spend figure that is actually connected to your business model. If your target CPA is £60 and you need 40 new customers per month, your floor is £2,400 per month before any platform learning-phase overhead.
If you do not yet know your CPA or CPL, that is the first problem to solve. Running advertising without knowing your cost of customer acquisition is like running a business without knowing your margin.
AdLeada insight
Among businesses that complete the AdLeada Ads Health Check, spend level is the second most commonly flagged dimension after tracking. A significant share of accounts spending under £1,000 per month show "Learning limited" status in their ad platforms, indicating their budget is below the algorithm's minimum viable threshold.
The SMB playbook: £1,000 to £10,000 per month
Focus over spread
At under £3,000 per month total, spreading budget across multiple platforms puts every channel below its learning-phase floor. The right approach:
- Focus on one channel where you have evidence your customers are active
- Run one campaign and one ad set until the learning phase exits (typically two weeks)
- Only add a second channel once the first is consistently profitable over at least two months
- Test creative variations within existing campaigns, not across new campaigns
- At £3,000 to £10,000 per month, you can properly run two channels simultaneously
Most small UK businesses get better results from £1,000 focused on one platform than from £3,000 split across three.
The mid-size playbook: £10,000 to £50,000 per month
Mix and attribution matter now
At this spend level you have enough budget to run two or three channels properly and to begin meaningful creative testing. The central question shifts from "can we afford to advertise?" to "where does each pound work hardest?"
- Allocate a minimum of 15% to testing new channels or audiences (keeping 85% on proven activity)
- Set a clear attribution model before adding channels, or you will not know what is driving growth
- Consider channel sequencing: awareness channels (audio, display, outdoor) feeding performance channels (Meta, Google) over time
- At £30,000 or more per month, commissioning a formal media mix review is worth the cost of an independent assessment
How to know if your current spend is actually working
Budget size is not the same as budget effectiveness. These four questions determine whether your current spend is working or just spending:
- What is your current cost per acquisition or cost per lead? If you do not know this number, your tracking is not complete.
- Is your ROAS above 1.5x as a minimum? Below 1.5x you are spending more than you are recovering unless you have strong customer lifetime value.
- Have your campaigns exited the learning phase? "Learning limited" status means the platform is not yet optimising and results will be inconsistent.
- What share of conversions can you attribute to paid advertising? If most conversions are "Direct" or "Unattributed", your attribution model needs work before you can make meaningful budget decisions.
If you cannot answer all four, you need better measurement before you need a bigger budget. The AdLeada Ads Health Check assesses your full advertising setup across five dimensions and tells you where the gap is.
Is your current spend actually working?
The AdLeada Ads Health Check takes two minutes and gives you an instant score out of 100 across your channel mix, spend level, tracking, brand visibility, and growth readiness.
Take the free Ads Health CheckNo account required. Results on screen immediately.
What the research says about ad spend and effectiveness
Several consistent findings from platform research are relevant to UK budget decisions:
- Google's internal research shows that advertisers on Google Search who bid below the first-page minimum typically receive very few impressions regardless of daily budget set.
- Meta's own data shows that ad sets spending below the learning-phase threshold have performance variance three to five times higher than those that have completed learning.
- UK average CPCs on Google Search have risen approximately 15 to 20% over the past two years across most sectors, meaning the effective budget floor has also risen.
- LinkedIn's B2B cost per lead in the UK is typically five to seven times higher than Meta, making it cost-prohibitive for most businesses spending under £5,000 per month total.
Score your advertising setup in two minutes
Six questions. Instant score out of 100. Tells you whether your spend level is working for you and where your advertising is strongest and weakest.
Start the Health CheckFree. No sales call.
Frequently asked questions
What percentage of revenue should I spend on advertising in the UK?
The percentage-of-revenue rule is a rough proxy, not a strategy. Typical UK ranges are 5 to 12% for established consumer businesses, 15 to 25% for growth-stage companies, and 2 to 5% for B2B services with long sales cycles. A more reliable approach is to work backwards from your target cost per acquisition and required customer volume, rather than forwards from revenue. The right number is whatever produces a profitable customer at the volume you need, up to the point where incremental spend stops producing incremental results.
Can I start Meta ads with £500 a month?
You can run Meta ads at £500 per month, but most UK accounts at this spend level will not generate enough conversion events to exit the learning phase. You should expect inconsistent results and understand that the algorithm is not yet optimising. This is not necessarily a reason to avoid Meta at £500 if you are testing the channel, but you should not measure results as if the campaign is fully operational. Consider starting with a traffic or engagement objective at low spend to gather initial data, then scaling to a conversion objective once you have baseline performance information.
How much should a small UK business spend on Google Ads?
For most UK small businesses, the minimum for Google Search to generate consistent results is £600 to £800 per month per campaign. Below this, you may not appear frequently enough to generate meaningful data. Google Shopping can sometimes work at lower budgets (from around £400 to £500 per month) because CPCs are typically lower. Performance Max requires at least £1,000 per month and an existing conversion history to function effectively. If your total advertising budget is under £1,000 per month, consider focusing on one Google campaign type rather than splitting across multiple.
Is it better to spend more on Google or Meta?
Google captures demand that already exists: people actively searching for what you offer. Meta creates demand: it shows your brand to people who were not actively looking. Neither is inherently better. The right answer depends on your product and funnel. If strong search volume exists for your category, Google Search typically performs well from the start. If you are building a new category, or your customers do not yet know to search for what you offer, Meta or awareness channels are more appropriate. Most businesses above £3,000 per month benefit from running both. See our guide on building the right media mix for a channel comparison.
How do I know when to increase my ad budget?
Increase budget when three conditions are true simultaneously: your campaigns have exited the learning phase, your cost per acquisition or lead is profitable, and your current budget is limiting delivery (shown by low impression share or frequent budget-limit notifications). If any one of these is not true, increasing budget will amplify the existing problem rather than solve it. If your CPA is unprofitable, more spend will produce more unprofitable acquisitions. Fix performance before scaling spend.
What is a good return on ad spend (ROAS) in the UK?
Break-even ROAS depends entirely on your margins. The formula is: break-even ROAS = 1 divided by your gross margin percentage. If your gross margin is 40%, your break-even ROAS is 2.5x. For e-commerce, target ROAS is typically 3 to 6x depending on margin structure. For lead generation, the equivalent metric is cost per lead benchmarked against your average lead-to-sale conversion rate and deal value. A "good" ROAS does not exist in isolation, only relative to your unit economics. See our guide to measuring whether your advertising is working for the full set of metrics to track.