Leveraging A/B Testing in Algorithms
Most small business owners I speak to in Derbyshire and across the East Midlands are spending money on marketing without really knowing what is working. They are running ads, posting on social media, sending emails, and updating their website based on gut feel. And gut feel, in my experience, costs money.
I have been doing this for 18 years. The businesses that consistently grow are not the ones with the biggest budgets. They are the ones that test before they spend. A/B testing is one of the most practical and underused tools available to any SMB owner who wants to make better marketing decisions without paying for guesswork.
In this post I am going to explain what A/B testing actually means for a small business, how it connects to the way algorithms decide who sees your content, and how you can start doing it without a data science team or enterprise software. By the end, you will have a clear framework for testing your marketing in a way that produces real, measurable improvements.
Free tool
Not sure how your marketing measures up right now?
Use the free BIG12 Scorecard to benchmark your current marketing across 12 key pillars and find out where you are leaving the most growth on the table.
Take the free scorecardWhat A/B testing actually means for a small business
A/B testing is the practice of running two versions of something at the same time, to the same type of audience, and measuring which one performs better. Version A is the control. Version B is the change. Everything else stays the same. The result tells you something real rather than something assumed.
It applies to almost everything in digital marketing. Subject lines in emails. Headlines on a landing page. The thumbnail on a YouTube video. The call to action on a Facebook ad. The opening line of a Google Business Profile post. The order of sections on a product page.
What makes it powerful is the discipline. When you test one thing at a time, you isolate the variable. When you isolate the variable, you learn something true. That knowledge compounds. Over time, a business that tests consistently will have a fundamentally better understanding of its audience than one that does not.
The difference between a test and a tweak
A lot of businesses change things without testing them. They update the homepage because someone on the team has a preference. They change the email subject line because last week's open rate was low. They swap the image in an ad because it feels a bit tired. None of that is testing. It is tinkering.
A proper A/B test defines what you are trying to improve before you run it, sets a sample size large enough to get a statistically meaningful result, runs both versions simultaneously, and measures against a specific metric. Without those four elements, you are not testing. You are guessing with extra steps.
How algorithms use testing signals to determine your reach
Platform algorithms on Google, Meta, LinkedIn, and email platforms are themselves running constant experiments. They are assessing how audiences respond to your content and using those signals to decide how much distribution to give it.
When you publish a post on LinkedIn, the algorithm initially shows it to a small sample of your connections. If those people engage with it quickly, the distribution expands. If they scroll past it, it stagnates. The algorithm is, in effect, running an A/B test on your content against everything else in the feed. Your post is competing for attention, and the signal it generates in the first hour often determines how far it travels.
The algorithm does not care what you intended. It measures what your audience actually did. If you are not testing and refining before you publish, you are sending work into that environment without knowing how it will perform.
Stuart Baddiley, Optimise Your Marketing
The same logic applies to paid advertising on Google and Meta. Both platforms use machine learning to optimise delivery of your ads. They rotate your creative variants and allocate more spend to the ones getting the best response rate relative to your objective. If you only give them one version of an ad, the platform has nothing to test. You are removing the mechanism that improves performance over time.
Understanding this means you should be designing your content with variation built in from the start. Not two completely different ads, but controlled variation. Same offer, two different headlines. Same post, two different opening lines. Let the platform and your own analysis tell you which works.
Where to run your first A/B tests as a Derbyshire SMB
The best place to start is wherever you are already spending time or money. There is no point running tests on channels you barely use. Pick one channel where you have a meaningful volume of activity and start there.
Email subject lines
Email remains one of the highest-return channels for SMBs. Most email platforms, including Mailchimp and Klaviyo, have built-in A/B testing for subject lines. You do not need to set anything up from scratch. Set a test, split your list, run both versions to a portion of your list, wait four hours, and send the winner to the rest. This is a twenty-minute job that can meaningfully improve your open rate over time.
A 15 to 20 percent improvement in open rate from subject line testing is not unusual once you have run four or five tests and started to understand your audience's language patterns. That is not a marginal gain. At scale, it changes your lead generation results significantly.
Landing page headlines
Your website homepage or key landing pages are the highest-leverage places to test copy. The headline is the first thing a visitor reads. It determines, in most cases, whether they stay or leave. A tool like Google Optimise (now sunset), VWO, or even a simple Squarespace page duplication can let you run a basic split test on your headline without touching the rest of the page.
I worked with a professional services firm in Derby that changed a single headline from "Experienced Accountants for Your Business" to "We Help Derbyshire Businesses Keep More of What They Earn." Enquiries from organic traffic increased by 31 percent over the following six weeks. Same page, same offer, same traffic source. One change. One test.
Ad creative variants
If you are running paid social or Google ads, you should always be running at least two creative variants in every campaign. Meta and Google will both tell you which is performing better against your objective. You should check this weekly, pause the underperformer, replace it with a new challenger, and repeat. This is the most basic form of test and measure that any business running paid ads should be doing as standard.
Client result
Consistent testing changed the trajectory
A trades business in the East Midlands we worked with had been running the same Google ad copy for over a year. After introducing a structured A/B testing schedule across their ad headlines and landing page, cost per lead dropped by 28 percent within three months. The work was not in writing better ads from scratch. It was in measuring which language their actual customers responded to.
See how we approach testingThe testing mistakes that waste time and money
I have seen businesses run tests badly and conclude they do not work. Usually the test itself was the problem.
Testing too many variables at once
If you change the headline, the image, the button colour, and the layout all at the same time, you have no idea which change caused any movement in your results. Test one element. Wait until you have a result. Then move to the next element. Discipline here is everything.
Ending the test too early
Small businesses often stop a test after a few days because one version looks like it is winning. Statistical significance takes time and volume. A sample of 40 clicks is not a reliable data set. Most tests need at least 200 to 300 interactions per variant before you can draw a confident conclusion. If your traffic volume is low, you may need to run the test for two to four weeks.
Testing without a clear objective
Define the metric before you run the test. Are you measuring open rate, click-through rate, time on page, or conversion? Testing without a pre-defined success metric means you will find whatever number looks best after the fact and call it a win. That is not testing. That is confirmation bias.
Ignoring external factors
Do not run a test across a bank holiday, a school half term, or a period when your business is unusually busy or quiet. Seasonal and contextual factors will distort your results. Keep your test periods as clean and consistent as possible.
Building a testing habit into your marketing
The businesses that get the most from A/B testing are not running one test a quarter. They have built testing into how they work. Every email campaign includes a subject line test. Every paid ad campaign runs two creative variants. Every major page update is followed by a measurement period with a control to compare against.
This does not require a large team. It requires a small amount of structure and the discipline to measure before you change. At OYM, we work with clients across Derbyshire and the East Midlands to build this habit systematically, connecting it to their broader BIG12 framework so that testing informs every pillar of their marketing, not just one channel.
Keep a simple testing log
You do not need specialist software to track your tests. A shared spreadsheet works. Record what you tested, when you ran it, what the hypothesis was, what the result was, and what you decided to do as a result. After six months, that log becomes an invaluable record of what your specific audience responds to. That knowledge belongs to your business. No agency or algorithm can take it away.
Let your data challenge your assumptions
Most business owners I meet have strong opinions about what their customers want to hear. The honest truth is that those opinions are often wrong, not because the owner does not know their customers, but because they are too close to it. A/B testing removes the opinion from the equation. It replaces "I think this will work" with "the data shows this works." That shift in mindset is worth more than any single test result.
How A/B testing connects to the BIG12 framework
Within the BIG12, test and measure is one of the twelve pillars precisely because it underpins everything else. You can have a good brand, strong SEO, an active social media presence, and a well-maintained CRM. But if you are not measuring how those elements perform and testing ways to improve them, you are not growing. You are maintaining.
A/B testing is the mechanism that turns your marketing activities into a learning system. Each test produces a result. Each result informs the next decision. Over time, this compounds into a meaningful competitive advantage, particularly in local markets like Derbyshire and the East Midlands where many competitors are not testing at all.
Free tool
Not sure how your marketing measures up right now?
Use the free BIG12 Scorecard to see where your test and measure approach stacks up against all 12 pillars of a strong marketing strategy.
Take the free scorecardThe challenge is never learning. It is doing.
Most business owners understand testing in principle. The gap is implementation. It is easy to read about A/B testing, agree it makes sense, and then continue running campaigns exactly as before because the day-to-day pressure of running a business leaves no room for structured experimentation.
I have seen this pattern across hundreds of businesses in Derbyshire and the wider East Midlands over 18 years. The knowledge is available. The tools are free or low-cost. What is missing is the habit, the structure, and in many cases, a second set of eyes to hold the process accountable.
That is exactly what we do at OYM. If you want to move from guessing to knowing, the starting point is understanding what your marketing looks like right now and where the biggest testing opportunities are. Let's have that conversation.
Book a free audit
Book a free 90-minute audit with Stuart
We will look at your current marketing, benchmark it against the BIG12, and give you a practical set of actions to take. No sales pitch. No fluff. Just 18 years of honest advice applied to your business.
Book your free audit