Marketing Attribution Models: First-Touch vs Data-Driven

Run the same campaign for a month, then open two dashboards. Google Ads says it drove 40 enrolments. Your analytics tool says 22. Finance, looking at the bank, counts 31. Nobody is lying. They are each using a different marketing attribution model — a rule for deciding which click, view or message gets the credit when a customer finally buys. Change the rule and the winner changes with it. This guide walks one real-looking buyer journey through the five models every marketer meets, shows exactly how the credit moves each time, explains what Google Analytics 4 actually offers today, and gives you a simple way to pick one and stop arguing with your own reports.

What a marketing attribution model actually does

A customer almost never buys on first contact. They find you through a Google search, watch a YouTube review a week later, message you on WhatsApp, then come back through a branded search and enrol. That is four touchpoints for one sale. Attribution is the question: who gets the credit for that enrolment — the first touch, the last, or everyone in between?

An attribution model is just the formula that answers it. It takes 100% of the credit for a conversion and divides it across the touchpoints the customer actually used. A "first-touch" model hands all 100% to the Google search. A "last-touch" model hands it all to the branded search. A "linear" model splits it evenly. None of them is objectively correct, because credit is a judgement, not a measurement.

This matters because the model silently decides where your next rupee goes. If your reports run on last-click, your awareness content looks worthless and you cut it — even though it started every journey. Understanding attribution is really about understanding your own marketing funnel and refusing to reward only its last step. If you want this foundation built properly rather than pieced together from scattered videos, a structured digital marketing course compresses years of trial and error into a few guided weeks.

Five attribution models, one buyer journey

Let us use a single journey the whole way through, so you can watch the credit move. Imagine a working professional deciding to upskill. Their path to enrolling looks like this:

Four touches, one enrolment worth 100 credit points. Here is how each of the five classic models splits those 100 points across the same journey.

First-touch and last-touch: the all-or-nothing pair

First-touch gives every point to the Google search that started it all. It flatters your top-of-funnel — your SEO, your awareness ads — and completely ignores whatever closed the deal. Last-touch (also called last-click) does the opposite: all 100 points go to the branded search at the end. It is the most common default in the world, and the most misleading, because branded search is usually just the customer typing your name after something else convinced them.

Linear, time-decay and data-driven: the shared-credit trio

Linear splits the 100 points evenly — 25 each — treating every touch as equally important. Time-decay gives more weight to touches closer to the sale, on the logic that recent nudges matter more. Data-driven is the modern approach: instead of a fixed rule, it uses machine learning to compare thousands of journeys that converted against ones that did not, then assigns fractional credit based on what actually seems to move people.

You will also meet a sixth classic model: position-based, sometimes called U-shaped. It is a compromise between the all-or-nothing pair — it hands 40% of the credit to the first touch, 40% to the last, and spreads the remaining 20% across everything in the middle. The idea is that the touch that found the customer and the touch that closed them both deserve the lion’s share, while the nurturing steps in between still earn something. It is intuitive, which is why it stayed popular for years, but it is still a fixed guess rather than a reading of your actual data.

Read the table top to bottom and one thing jumps out: the YouTube review is worth either 0 or 25 credit points depending purely on the model you picked — nothing about the customer changed. This is the single most important idea in attribution. The chart below isolates the first touch, your Google search, to show how violently its credit swings.

How attribution works in Google Analytics 4 today

Here is where theory meets the tool most Indian marketers actually use. For years, Google Analytics let you choose from a long menu of models. That changed. Google retired the four rules-based models — first-click, linear, time-decay and position-based — from Analytics, removing them for new conversion actions from May 2023 and completing the switch by September 2023. If you learned attribution from an older course, half the models you studied no longer exist as options in the interface.

Today GA4 gives you exactly three choices: data-driven attribution (now the default), paid and organic last click, and Google paid channels last click. Data-driven is the model Google pushes everyone toward, and for most accounts it is the right call — it considers up to the last 50 interactions in a path and reads non-converting journeys too, which no fixed rule can do. You change the model under Admin, in the attribution settings, and it reshapes the conversion credit across your reports. If the mechanics of the interface itself are new to you, our walkthrough of the 5 GA4 reports that matter is the fastest way to get oriented.

One caution: data-driven attribution needs volume to work. If your property records only a handful of conversions a month, the model has too little signal, and you may be better served by a simple, consistent last-click view you fully understand than by a black box you cannot interrogate.

Which attribution model should you choose?

For a small or mid-sized budget, the honest answer is not "the most sophisticated one." It is "one you understand, applied consistently." Switching models every quarter guarantees your trend lines are meaningless. Use this table to match the model to your situation.

The attribution mistakes that quietly waste budget

Most attribution damage is self-inflicted. These are the errors we see most often from otherwise sharp marketing teams:

• Trusting last-click by default. It is the path of least resistance and it systematically punishes the awareness work that fills the top of your funnel. Teams cut the very campaigns that started every journey.
• Comparing numbers across tools. Google Ads, GA4 and your CRM each attribute differently and use different windows. Lining them up side by side and expecting a match wastes hours. Pick one source of truth for each decision.
• Switching models mid-year. Every switch breaks your trend. If you must change, re-baseline and tell everyone the old and new numbers are not comparable.
• Forgetting the offline touch. In India, a WhatsApp chat or a phone call often closes the sale, and no model sees it unless you feed that event back in. Your dashboard’s "direct" and "branded" conversions are often hiding this human step.
• Confusing attribution with the full picture. Attribution tells you which touch got credit, not whether a lead was any good. Pair it with lead quality — the difference between an MQL and an SQL — before you judge a channel.

Notice that none of these fixes needs a bigger budget or fancier software. They need discipline and a shared definition of what a conversion is worth.

What to do next

Attribution is not a tool you install; it is a decision you make and then defend consistently. Start here: open GA4, confirm whether you are on data-driven or last-click, and simply know which one your reports run on — most teams do not. Then pick the single model that matches your traffic volume and sales cycle, write it down, and judge every channel through that one lens for a full quarter before you touch it again.

From there, connect attribution to the metrics that pay the bills. A model that tells you a channel gets credit is only useful next to its cost — so read it alongside the marketing metrics that actually matter like CPC, ROAS and CAC. Credit without cost is vanity; credit against cost is strategy. That shift — from counting conversions to understanding what earned them — is what separates a marketer who guesses from one who decides.

Frequently Asked Questions

What is the difference between first-touch and last-touch attribution?

First-touch attribution gives 100% of the credit for a conversion to the very first interaction a customer had with you — typically a discovery channel like search or an awareness ad. Last-touch gives all the credit to the final interaction before they converted, which is often a branded search. First-touch tells you what creates demand; last-touch tells you what closes it. Neither tells the whole story alone.

Which attribution model does Google Analytics 4 use by default?

GA4 uses data-driven attribution as its default reporting model. Google retired the older first-click, linear, time-decay and position-based models through 2023, so GA4 now offers only three choices: data-driven, paid and organic last click, and Google paid channels last click. You can change the model in the Admin attribution settings, and it reshapes conversion credit across your reports.

Is data-driven attribution better than last-click?

For accounts with steady conversion volume, data-driven attribution is usually more accurate because it learns from real converting and non-converting journeys instead of following a fixed rule. But it needs enough data to be reliable and is hard to audit. If you record only a few conversions a month, a simple last-click view you fully understand can be the more honest choice.

What is an attribution lookback window?

The lookback window is how far back a model looks to include touchpoints before a conversion. In GA4, the default window for most conversions is 90 days, so an interaction older than that gets no credit. Choosing a window that matches your real sales cycle matters: too short and you miss early discovery touches, too long and you credit interactions that had no influence.

Which attribution model should a small business in India use?

Pick one model you understand and apply it consistently for at least a quarter. For most small budgets, data-driven attribution in GA4 is a sensible default if you have steady traffic; otherwise, a clear last-click view works. The bigger win is feeding offline touches — WhatsApp chats and phone calls that often close the sale — back into your measurement so they stop hiding as "direct" conversions.