Analytics
Marketing attribution for small businesses: what to do when every platform claims the sale
Add up the sales each platform claims and you will often get more than you actually made. You don't need a data team to fix this, just a few habits and three numbers.
By the ARIQ Editorial Team 4 min read
A furniture shop spends on Google Ads, Meta ads and an email newsletter. At the end of the month Google claims 40 sales, Meta claims 35, the email tool claims 20. The shop made 60 sales in total. Nobody is lying; each platform counts every sale it touched, and most customers touched more than one.
Attribution is the question of how to share the credit. Big companies spend fortunes on it. Small businesses need something simpler that still leads to good decisions.
Why the numbers overlap
A typical path to purchase: someone sees an Instagram ad on Tuesday, searches your brand on Google on Thursday, clicks a search ad, browses, leaves, then buys on Sunday from your newsletter. Meta counts the sale because of Tuesday's view. Google counts it because of Thursday's click. The email tool counts it because of Sunday's click. Your store records one order.
Each platform also uses its own attribution window and its own rules about views versus clicks, and privacy settings hide part of the journey from all of them. Our article on Meta ads reporting goes into the Meta side in detail.
The attribution models worth knowing
| Model | How credit is given | Good for |
|---|---|---|
| Last click | All credit to the last thing clicked before buying | Simple, but undervalues awareness channels |
| First click | All credit to the first touch | Seeing what introduces new customers |
| Linear | Credit split equally across touches | A rough, fair-looking split |
| Data-driven | Credit shared by a statistical model | Accounts with plenty of conversions; the GA4 default |
| Platform self-reporting | Each platform counts every sale it touched | Comparing ads within one platform, not across platforms |
No model is “true”. They are different lenses. The mistake is mixing them: comparing Meta's self-reported numbers with Google Analytics' data-driven numbers and drawing conclusions from the difference.
A simple system that works
1. Tag every link you control
Add UTM parameters to every link in ads, emails, social posts and partner sites: utm_source, utm_medium and utm_campaign at minimum. Use a consistent naming scheme written down in one place. Untagged links show up as “direct” or “unassigned” in Google Analytics, and that hides what is working.
2. Use one source of truth for totals
Revenue and order totals come from your store, CRM or accounts, never from an ad platform. That number is what you can bank.
3. Watch three numbers together
- Blended ROAS (or MER): total revenue divided by total ad spend. It cannot be inflated by overlapping attribution.
- Cost per new customer: total marketing spend divided by first-time customers. Growth that only comes from repeat buyers is not growth from marketing.
- Channel trends in one tool: use Google Analytics, with its own consistent model, to compare channels against each other over time.
4. Ask customers
Add a one-question survey after checkout or on the enquiry form: “How did you first hear about us?” It is not precise, but it catches word of mouth, podcasts, YouTube and other channels that no tracking pixel sees.
5. Test, occasionally
When you are unsure whether a channel is pulling its weight, change its spend noticeably for a couple of weeks and watch the totals from your source of truth. If sales barely move when you halve a channel, it was claiming more than it caused.
What to report to a client or a boss
Lead with total revenue or leads and blended return, then show each channel's trend with a clear note about which model the number comes from. It is honest and it stops the monthly argument about whose numbers are right. Our monthly client report template shows where this fits.
See every source side by side
The hardest part of attribution for a small team is simply getting the numbers in one place. ARIQ connects Shopify, Meta, Google Ads and GA4, calculates blended ROAS and cost per result automatically, and labels where each number comes from. If you run a store, start with the Shopify metrics that decide growth, then try ARIQ free.
Frequently asked questions
What is the best attribution model for a small business?
There is no single best model. Use your store or CRM for totals, Google Analytics for comparing channels over time, ad platforms for comparing ads within that platform, and blended return on ad spend to judge overall efficiency.
What are UTM parameters?
UTM parameters are tags added to the end of a link, such as utm_source=newsletter, that tell analytics tools where a visit came from. They make paid, email and social traffic visible in Google Analytics.
What is blended ROAS?
Blended ROAS, also called MER, is total revenue divided by total advertising spend across all channels. It ignores which platform claims each sale, so it cannot be inflated by overlapping attribution.