Summary

Customers rarely see one ad and buy. They discover you on social, return via search, read an email, and finally convert, and the question of which of those touchpoints gets credit for the sale is the problem attribution solves. Get it wrong, and you cut the channels that quietly assist conversions while overspending on the ones that merely close them. This guide explains attribution in plain language, walks through the main models with South African examples, exposes the trap of last-click thinking, and helps you choose an approach that reflects how your customers actually buy.

What attribution is and why it is hard

Marketing attribution is the practice of assigning credit for a conversion, a sale, a lead, an enquiry, to the marketing touchpoints that contributed to it. In a world where a customer interacted with only one piece of marketing before buying, attribution would be trivial: that one thing gets the credit. But that is almost never how buying works.

A realistic South African customer journey might look like this: a person first encounters your business through an Instagram ad, does nothing immediately, later searches your brand name on Google and visits your site, leaves again, receives an email a week later, clicks through, and finally returns via a Google search and buys. Five touchpoints across several channels and a couple of weeks. Which one earned the sale?

This is the attribution problem. Each touchpoint played a role, the Instagram ad created awareness, the email re-engaged, the final search closed, but most simple measurement gives all the credit to just one of them, usually the last. That distortion matters, because you make budget decisions based on which channels appear to drive sales, and if your measurement systematically misattributes credit, you systematically misallocate budget.

The core problem: Customers touch several channels before buying. How you split the credit among them decides which marketing looks successful, and therefore where you spend, so attribution is not academic; it is a budget decision.

First-click attribution

First-click (or first-touch) attribution gives all the credit for a conversion to the first touchpoint in the customer's journey, the channel that first brought them into contact with your business.

In our example, first-click would credit the Instagram ad with the entire sale, on the logic that without that first introduction, none of the rest would have happened. The strength of this model is that it highlights the channels that are good at discovery and awareness, the ones filling the top of your funnel with new prospects. Its weakness is that it ignores everything that happened afterwards, giving no credit to the email and search that actually moved the customer toward buying. It can make awareness channels look like the whole story when they are really just the opening chapter.

First-click is useful when you specifically want to understand which channels are best at introducing new people to your business, which matters a great deal if your challenge is generating awareness and new demand rather than converting existing interest.

Last-click attribution

Last-click (or last-touch) attribution gives all the credit to the final touchpoint before conversion, the channel the customer used immediately before buying. This is the most common model, often because it is the default in many analytics setups, and that ubiquity makes it the most dangerous.

In our example, last-click credits the final Google search with the whole sale, ignoring the Instagram ad that created the awareness and the email that re-engaged the customer. The appeal is simplicity: it is easy to measure and points to what closed the deal. But the trap is severe. Last-click systematically overcredits channels that appear at the end of journeys, typically branded search and direct, while undercrediting the awareness and nurturing channels that created and sustained the demand in the first place.

The real-world consequence is brutal and common: a business looks at last-click data, sees that its social or display advertising 'drives few sales', and cuts it, only to watch overall sales decline, because that awareness channel was quietly feeding the top of the funnel that the closing channels then harvested. Last-click made the assisting channel invisible, so it got cut, and the whole funnel suffered. Understanding this trap is perhaps the single most valuable thing about learning attribution.

Linear and position-based models

Between the extremes of crediting only the first or only the last touchpoint sit models that share credit across the journey, which better reflect reality.

Linear attribution gives equal credit to every touchpoint in the journey. In our example, each of the five touchpoints would receive a fifth of the credit. Its strength is that it acknowledges every touchpoint contributed something, which corrects the blindness of single-touch models. Its weakness is that it treats every touchpoint as equally important, which is rarely true; a passing ad impression and a final decisive search did not contribute equally.

Position-based (often called U-shaped) attribution gives more credit to the first and last touchpoints, treating the introduction and the close as most important, while sharing the remainder among the middle touchpoints. This reflects a common intuition: the channel that first brought the customer in and the channel that closed the sale deserve significant credit, while the nurturing touchpoints in between deserve some but less. For many businesses, position-based attribution offers a sensible balance, recognising both discovery and closing while still acknowledging the middle.

There are more sophisticated approaches still, including data-driven models that use your actual conversion data to assign credit based on each touchpoint's observed contribution. These can be powerful but require sufficient data to be reliable. For most SMEs, understanding the main models and choosing one that reflects their reality is more practical than chasing a complex model.

Choosing a model for your business

There is no single correct model; the right one depends on your business, your sales journey and what you are trying to learn. The key is to choose deliberately rather than defaulting to last-click without realising it.

If your priority is understanding which channels introduce new prospects, lean toward first-click insight. If you want to understand the full journey and value your assisting channels fairly, a linear or position-based model serves better. Many businesses benefit from looking at more than one model: comparing how channels perform under different attribution lenses reveals which are good at starting journeys, which at closing them, and which quietly assist throughout. A channel that looks weak under last-click but strong under first-click is an awareness engine you would be foolish to cut.

The practical goal is to avoid the costly mistake of judging every channel by last-click alone. Even simply being aware that your default reporting is probably last-click, and mentally adjusting for the assisting channels it undercredits, puts you ahead of most businesses making budget decisions on distorted data.

Making attribution practical

Attribution can become an academic rabbit hole, so keep it grounded in better decisions. The aim is not perfect measurement, which is genuinely difficult given privacy changes, multiple devices and offline touchpoints, but better-informed budget allocation than naive single-touch thinking allows.

Practically, ensure your analytics is set up to capture the channels driving your traffic, use consistent campaign tagging so you can see how channels work together, and review your data through more than one attribution lens before making cut-or-invest decisions. Pay particular attention to assisted conversions, the touchpoints that contributed to journeys without being the final click, since these reveal the channels that last-click hides.

Above all, treat your numbers with appropriate humility. Attribution is a model of reality, not reality itself, and every model simplifies. The goal is to use it to ask better questions, which channels are really contributing, which look good only because they close pre-existing demand, where is the awareness coming from that the closing channels harvest, rather than to find a single magic number. A South African business that understands attribution well enough to avoid cutting its quiet assisting channels, and to fund the awareness that feeds its funnel, is already making far better decisions than competitors spending on last-click instinct alone.

Related Juicy Designs resources

• Analytics and reporting services
• What is marketing attribution? (glossary)
• GA4 for beginners: a practical SA setup guide
• Marketing channels compared: where to spend first
• Marketing strategy services

Frequently asked questions

What is marketing attribution?

Marketing attribution is how you assign credit for a sale or lead to the marketing touchpoints that influenced it. Because customers usually interact with several channels before buying, attribution determines which marketing gets the credit, which in turn shapes where you spend your budget.

What are the main attribution models?

The main models are first-click (all credit to the first touchpoint), last-click (all credit to the final touchpoint), linear (equal credit to every touchpoint), and position-based or U-shaped (more credit to the first and last, less to the middle). Data-driven models also exist for businesses with enough data.

Why is last-click attribution a problem?

Last-click gives all credit to the final touchpoint, systematically overcrediting closing channels like branded search while ignoring the awareness and nurturing channels that created the demand. This leads businesses to cut assisting channels, then watch overall sales fall because the funnel lost its top.

Which attribution model should I use?

There is no single correct model; choose deliberately based on what you want to learn. First-click highlights discovery channels, linear and position-based value the full journey more fairly. Looking through more than one model reveals which channels start, assist and close journeys.

What is an assisted conversion?

An assisted conversion is one where a touchpoint contributed to the customer's journey without being the final click that completed it. Assisted conversions reveal the channels that last-click attribution hides, which is why reviewing them protects you from cutting valuable assisting channels.

Can I measure attribution perfectly?

No. Privacy changes, multiple devices and offline touchpoints make perfect attribution genuinely difficult. The realistic goal is better-informed budget decisions than naive single-touch thinking allows, by using consistent tracking, reviewing multiple models and treating the numbers as a useful model rather than exact truth.