Which channel gets the credit: attribution for budget decisions

Fibvo editorial · 28 April 2026 · 5 min read

Google Ads says it brought in most of last month's enquiries. Meta says much the same. The SEO report claims a good share as well. Add the three together and the total is higher than the number of enquiries your sales team actually handled. Nobody is inventing numbers. Each tool is answering a narrower question than the one you are asking, which is where the next rupee of budget should go.

Why every platform claims the same enquiry

Attribution is the rule that decides which touchpoint gets credit for a result. Each advertising platform applies its own rule to its own data, and each one sees only itself.

Picture a modular kitchen company working in three cities. A homeowner sees an Instagram ad one evening and scrolls past. Two days later they search for the company by name, click a Google ad and browse the gallery. The following week they type the web address directly and submit the enquiry form. Meta can count that enquiry, because its standard attribution setting includes people who viewed an ad without clicking. Google Ads counts it, because an ad click fell within its conversion window. GA4 never saw the Instagram view, so it credits the Google ad. Three reports, one kitchen.

So the question of which channel gets the credit has no single correct answer. The question that helps with budget is a different one.

The useful question is not which channel touched the sale, but what you would lose if that channel went quiet.

Read GA4's attribution for what it is

GA4 offers data-driven attribution, which spreads credit across the touchpoints it observed by comparing journeys that led to a key event with those that did not, and last click models, which give all the credit to the final non-direct channel. Google has retired the older rule-based models, such as first click and linear. You can choose the reporting model and the lookback window in the property's attribution settings.

The conversion paths report is worth a regular look. It shows the sequence of channels before key events, so you can see whether a channel tends to open journeys or close them. A channel that keeps appearing early and rarely last is doing work that last click reporting hides.

Be clear about what GA4 cannot see: ads viewed without a click, a recommendation from an architect, a conversation at a home expo, a link forwarded by a neighbour. Consent choices and browser privacy features break some journeys too. GA4 attribution describes the clickable part of the journey only.

Check platform claims against what customers tell you

Platforms report what they tracked. Customers tell you what they remember. Each month, put the source your sales team recorded for each enquiry next to the channel the platforms credited, and look at where they disagree. If enquiries credited to brand search keep saying 'saw you on Instagram' or 'my interior designer suggested you', the brand search ad is collecting demand another channel created, and cutting Instagram would shrink those searches later.

People forget and simplify, so read the answers as patterns, not precise records. Our post on measuring cost per enquiry sets out how to capture the source of every enquiry.

Test by turning a channel down

The strongest evidence comes from changing something on purpose and watching the total. For a business in several cities, a geographic holdout is practical.

• Pick two cities with similar enquiry history and a similar mix of channels.
• In one of them, pause or sharply reduce a single channel for at least four to six weeks. Leave everything else unchanged in both.
• Measure total enquiries from your CRM or sales register in each city, not the numbers the platforms report.
• Compare the test city with the control city over the same weeks, and with its own figures from the same period last year.

If enquiries in the test city barely move, the channel was mostly claiming credit for demand that would have come anyway. If they drop clearly, the channel is doing real work, whatever its dashboard suggests. Avoid test periods that cover Diwali, the financial year-end or any time your market behaves unusually. With low enquiry volumes, run the test longer rather than reading meaning into a short, noisy dip.

Turn the evidence into a budget split

With platform reports, customer answers and test results side by side, stop ranking channels on one number and group them by the job they do.

• Demand capture, such as brand search, high-intent search terms and retargeting. Judge these on cost per enquiry, and expect them to take credit for demand other channels created.
• Demand creation, such as social ads, video, content and events. Judge these on whether branded searches in Search Console and total enquiries rise while they run, and on what customers tell you.

Set a budget for each group, review it quarterly and move money in steps. Change one channel at a time so you can tell which change caused which result. When tools disagree, trust the total enquiry count in your own records over any single platform's report.

---

Written by Fibvo — Bengaluru and Erode, India — working with clients in other time zones. Questions, or think we have it wrong? Tell us.