MARKETING BASICS · 7 MIN READ

Online Marketing Is a Numbers Game: What New Business Owners Should Know

Online marketing can feel unpredictable when you are starting out. One article receives attention, another barely gets noticed, and a busy week does not necessarily produce a sale.

Numbers help you understand what is happening. They can show whether people are finding your content, taking the next step and buying. But calling marketing a numbers game does not mean publishing endlessly or treating people as clicks. Relevance, trust and a suitable offer still matter.

The useful approach is to measure a few clear steps, understand their limits and decide what to improve next.

🔎 Start with the customer journey

Map the route someone takes before becoming a customer. For an affiliate content site, that might be: discover an article, visit the page, click a relevant product link and complete a qualifying purchase with the seller.

Each step answers a different question. Impressions tell you about recorded exposure. Visits tell you people reached your site. Outbound clicks show interest in another page. Credited sales show transactions attributed to your affiliate account under the programme’s rules.

Do not mix them together. A social post shown 10,000 times has not necessarily brought 10,000 people to your website. The same person may also generate several impressions, visits or clicks.

Define your rates before comparing them

A percentage is only useful when you know what sits above and below the division line. For a simple page-level calculation, you might track sessions containing an affiliate click divided by article sessions, then multiply by 100.

Google Ads defines clickthrough rate as clicks divided by impressions. Its conversion-rate definition uses conversions divided by trackable ad interactions. Those definitions illustrate why a platform’s label may differ from the calculation in your own spreadsheet.

Write down the event, denominator, date range and data source for every metric. If the denominator is zero, record the rate as unavailable rather than inventing a percentage.

📊 Follow a fictional example

Imagine a month in which an article records 1,000 sessions. In 100 of those sessions, the visitor clicks an affiliate link once. The affiliate report records those 100 clicks and eventually credits two qualifying sales, each paying £30.

For this illustration, assume the reports align and the sales belong to those clicks. Real reporting is often less tidy.

These numbers describe the example only. They do not mean every 100 clicks will produce two sales. Later reversals could reduce the £60, and commission before costs is not profit.

Separate earnings from the cost of generating them

Record software, advertising, platform fees and other relevant operating costs. Track your working time separately too: unpaid hours still matter when deciding whether the business fits your life.

Suppose the fictional £60 commission falls to £30 after one £30 reversal. If relevant operating costs were £20, £10 remains before tax and before putting a value on your time. That is very different from celebrating £60 as take-home income.

For paid traffic, compare costs and retained earnings at the same stage. Earnings per affiliate click cannot be directly compared with the price of an ad click to your article: some article visitors never click through to the seller.

Our guide to high-ticket and low-ticket affiliate programmes explains why headline commission alone cannot settle which offer is suitable.

Look for the step that needs attention

If impressions rise but visits remain low, examine whether your title and preview explain a relevant benefit. If visitors arrive but rarely follow the product link, check the match between the article, reader and offer.

If outbound clicks occur without credited sales, investigate tracking, the seller’s page, pricing and audience fit. A reporting delay or attribution difference may also matter. Do not automatically conclude that the whole product is unsuitable.

Our article on choosing affiliate products can help you review fit. Improving the explanation should also respect the principles in recommending products ethically; pressure and exaggerated claims are not good fixes for weak results.

Ad: If you want to examine the site’s featured option, view its product page. Assess the current details and terms yourself. The hypothetical figures above do not describe that product’s performance.

🧭 Give small samples the caution they deserve

One sale from ten clicks produces a recorded rate of 10%. If the next ten clicks produce no sales, the combined rate becomes 5%. Nothing about the product necessarily changed; the sample is simply small.

There is no universal number of clicks that proves an offer works. The amount of evidence needed depends on the underlying rate, the change you are testing and the uncertainty you can tolerate. Early figures are signals for investigation, not a basis for confident income forecasts.

Compare similar traffic sources and allow time for delayed purchases and adjustments. A visitor researching a topic casually is different from someone actively comparing purchase options.

Keep a short review routine

Record the period, traffic source, relevant sessions, outbound clicks, credited sales, retained commission, costs and one observation. Note what changed so you can interpret the next report.

Check that links and reporting work, then choose one useful improvement. At low traffic levels, clear usability fixes may be more practical than attempting to declare a winning A/B test. A before-and-after increase alone does not prove your edit caused it.

Use a manageable weekly content routine to review progress without refreshing dashboards all day.

Frequently asked questions

What is a good conversion rate?

There is no single rate that suits every business. Define the action and denominator, then consider audience intent, offer, traffic source and profitability.

Does more traffic always mean more sales?

No. Additional visitors may have different needs or little interest in buying. Relevant traffic and a clear offer matter alongside volume.

How often should I check results?

A weekly operational review is a reasonable starting routine, with longer comparisons when activity is low. Check sooner if a link or tracking system appears broken.

Should I buy ads to collect data faster?

Spending can produce clicks without useful customers. First understand the offer, tracking and costs. Set any experiment’s budget and stopping rules before spending; do not assume initial results will continue.