Insights
How long should you give a new marketing channel before judging it?
- Marketing strategy
- Budgeting
- Measurement
The short answer
A new marketing channel usually needs 8 to 12 weeks, and often 20 to 30 qualified enquiries, before a small business can judge it fairly, so judge early on cost per qualified enquiry and set a written stop rule before you start.
Most new marketing channels need 8 to 12 weeks, and often 20 to 30 qualified enquiries, before a small business can judge them fairly. If monthly spend is under £500 or the sales cycle is long, allow three to six months. A decision made after two quiet weeks is usually a reaction to noise, not evidence.
How long does a new marketing channel take to show results?
A new channel produces two different things: activity and evidence. Activity appears at once, in the form of impressions, clicks, calls and form starts. Evidence takes longer, because a click has to become a conversation, a quote, and sometimes a sale.
As a working guide, paid search can show whether people click within two to four weeks, but conversion evidence usually needs 8 to 12 weeks. Paid social often needs 6 to 10 weeks because the creative and audience need testing. Email to an existing list can show a pattern in two or three campaigns. Organic search is slower, typically three to six months before useful evidence, and longer in competitive markets.
These are ranges, not promises. The right test length depends on how many enquiries you need, how long your sales cycle runs, and how much you can spend without harming the business.
Why does a channel look bad in its first month?
Small numbers exaggerate everything. Suppose a campaign produces 500 clicks and 2 per cent of those clicks become enquiries. That is 10 enquiries. If one becomes a sale, the channel looks excellent. If none do, it looks dead. The underlying conversion rate has barely been measured either way.
The first month also carries setup costs that do not repeat. Tracking is corrected, the landing page is rewritten, the audience is narrowed, and the sales team learns which enquiries are worth pursuing. Those tasks consume budget and attention without producing a fair conversion rate.
Seasonality can distort the picture too. A four week test that lands in August or late December is not a fair sample for many businesses. Compare with a normal trading period, or extend the test to cover one.
How many enquiries do you need before deciding?
Start with your closing rate and your gross profit per sale. If you close about one enquiry in five, 20 qualified enquiries should produce around four sales. Twenty is still a thin sample, but it is enough to see whether the cost per enquiry is in the right area. Where volume allows, 30 to 50 qualified enquiries gives a steadier pattern.
Then compare cost per enquiry with what you can afford. If a sale produces £600 of gross profit and you close one in four, you can afford up to £150 per qualified enquiry before other costs. A channel charging £80 per qualified enquiry has room to work. One charging £220 does not, unless the close rate or average sale is higher than assumed.
This arithmetic is more useful than a generic return on ad spend target. It tells you what the channel has to achieve in your business, not in a textbook example.
What should you measure before the sales arrive?
Measure qualified enquiries, not raw leads. A form filled in by a student writing a school project is not the same as a form from a property manager with a budget and a date. The sales team should mark which enquiries were worth a conversation.
A simple scorecard has four lines: cost per qualified enquiry, qualified enquiry to sale rate, average gross profit per sale, and time from first enquiry to sale.
Track the source on every enquiry. Use consistent campaign tags on links, a call tracking number where calls matter, and a required source field in the customer record. Without that discipline, a channel that produces late sales will look worse than it is, because the sale gets credited to a later visit or a direct search.
When should you stop a channel early?
Stop before the test ends if the channel produces no qualified enquiries after a fair spend, or if cost per qualified enquiry is more than double your affordable ceiling for two consecutive months. Stop immediately if the ads appear against unsuitable content, if the offer breaks a compliance rule, or if the leads are consistently outside your service area.
There is also an operational reason to stop. If nobody can answer new enquiries within one working day, paid traffic will amplify a weakness rather than build a pipeline. Fix the response process first. A channel that sends ten good enquiries into an inbox nobody reads is not a marketing failure, it is an operations failure.
A written stop rule protects you from both panic and stubbornness. Decide the rule before the first pound is spent.
What is a fair test plan for a small budget?
A fair test does not need a large budget, but it needs enough room to produce a pattern. For a business spending £600 to £1,200 in total, a 12 week test is usually more informative than a three week burst.
- Choose one channel and one action, such as a booked call or a quote request.
- Set the test length at 8 to 12 weeks, or longer if the sales cycle runs beyond two months.
- Agree the budget in advance, including any setup fee.
- Tag every link and record the source on every enquiry.
- Check the account weekly for faults, but judge performance monthly.
- Review against the stop rule at the end, not after every quiet day.
If the budget is under £200 a month, do not split it across three platforms. One channel with enough data beats three channels with none. If the sales cycle is six months, judge the channel at 12 weeks on qualified opportunities, then review closed revenue at six months.
Is it ever right to judge a channel in the first week?
Yes, for technical faults. If the tracking is broken, the ads are disapproved, the landing page does not load, or the traffic is plainly from the wrong country, fix or pause it at once. That is not judging performance, it is checking that the test is running.
Do not judge the offer, the audience or the channel after one week. There is not enough information. The useful question at that stage is simple: is the machine switched on and are the right people arriving? The commercial verdict comes later.
Keep reading
What is Retrieval-Augmented Generation and do you need it?
Learn how Retrieval-Augmented Generation (RAG) works, what it costs in 2026, and whether your business needs it for internal or customer search.
Read the articleWhat does a core web vitals improvement cost a small business in 2026?
A realistic price range for fixing core web vitals issues in 2026, the tasks that drive the cost up or down, and when to skip the work.
Read the articleFind out how visible you actually are
A free written audit of how search engines and AI assistants currently read your website. No obligation, and no sales call required to receive it.
Prefer to talk first? Email contact@luiinteractive.com or message +44 7349 961542 on WhatsApp.