See which channels and campaigns drive revenue, and how far the numbers can be trusted.
Novaris brings ad platform data and store orders together so you can compare channels on one basis. It also shows where attribution is an estimate, because privacy rules and blocked tracking mean no tool sees every conversion.
What is included.
Revenue by channel and campaign
Orders and revenue attributed to each connected channel, campaign and ad, over the date range you choose.
Platform versus store view
What each ad platform reports next to what your store actually recorded, so gaps and double counting become visible.
Cross-channel paths
Where tracking allows it, the touchpoints a customer had across channels before buying, not just the last click.
Model and window settings
The attribution model and lookback window are stated on every report, and you can switch them to compare views.
How attribution works, and where it stops
Attribution is the attempt to give credit for a sale to the ads that led to it. Each ad platform does this for itself: Meta, TikTok and Google each count the purchases they believe they influenced, using their own models and windows. When a customer sees ads on two platforms before buying, both may claim the same order, which is why platform totals often add up to more than the store's real revenue.
Novaris reads results from your connected ad accounts and orders from your connected store, then lines them up on the same dates and currency. Purchase events from your pixels and tags, along with tracking parameters on ad links, help link an order to the campaign that brought the visit. The result is one table where platform-reported conversions and store orders can be compared side by side.
You can choose how credit is shared, for example to the last click, the first touch or spread across several touchpoints, and how far back to look. No model is the correct one. Each answers a different question, and comparing two of them is often more useful than trusting one. Reports always show which model and window they use, so numbers are not compared on different bases.
Attribution has real limits. Browser privacy features, ad blockers, consent choices and app tracking rules mean some visits and conversions cannot be tied to an ad at all. Platforms fill part of that gap with modelled conversions, which are estimates. Novaris labels what is measured and what is modelled where the source makes that clear, but it cannot recover data that was never collected.
What better attribution helps you decide
Where the next dollar goes
Budget decisions based on revenue by channel are more grounded than ones based on clicks or platform-reported totals alone.
Which campaigns to cut
A campaign that looks strong in its own ad manager but brings few store orders is easier to spot.
How channels work together
Some channels introduce customers and others close the sale. Path views help avoid cutting a channel that starts journeys.
A shared set of numbers
Founders, media buyers and accountants can look at the same report instead of arguing over different dashboards.
What to have ready
Attribution is only as good as the tracking behind it. These items have the biggest effect on data quality.
- Pixels and tags on every pageThe Meta Pixel, TikTok Pixel and Google tag installed on your store, firing on page view, add to cart, checkout and purchase.
- Server-side events where availableServer-side connections such as the Meta Conversions API can recover some events that browsers block. Your store app may offer them.
- A connected storeStore orders are the reference point. Connecting your store, such as Shopify, lets reports compare against real sales.
- Tracking parameters on ad linksUTM or similar parameters on every ad, with a consistent naming pattern, so visits can be traced to campaigns.
- A cookie consent setupA consent banner that matches the privacy laws of the countries you sell to. Consent choices affect what can be tracked.
- One currency and time zone decisionAgree which currency and time zone reports use, since ad accounts and stores are often set up differently.
From scattered numbers to one attribution view
- 1Day 1
Trial call
We look at your channels, how you judge performance today and which numbers you trust least.
- 2Day 1 to 2
Store and ad accounts connected
You authorize your store and each ad account. We check that purchase events reach each platform.
- 3Days 2 to 5
Tracking checked
We look for missing events, duplicated purchases and links without tracking parameters, and list what to fix.
- 4Usually within the first week
First comparison
Platform-reported results and store orders appear side by side for the same period, with the model and window shown.
- 5Weeks 2 to 4
Models compared
With more data, you compare two or three attribution views and agree which one guides budget decisions.
- 6Ongoing
Regular review
Tracking breaks when themes, apps or checkout change. A monthly check catches gaps before they distort reports.
Platform reports versus a combined view
| Manual | With Novaris | |
|---|---|---|
| Source of numbers | Each ad manager on its own | Connected ad accounts plus store orders |
| Double counting | Hard to see | Shown as the gap between platform totals and store revenue |
| Attribution model | Set separately on each platform | Chosen and shown on every report |
| Cross-channel journeys | Rarely visible | Shown where tracking data allows it |
| Modelled conversions | Mixed in without notice | Labelled where the source identifies them |
| Privacy limits | Apply | Still apply |
No attribution tool sees every conversion. Privacy rules, consent choices and blocked tracking leave gaps that can be estimated but not fully measured.
Reading attribution without over-trusting it
Treat attribution as a guide to direction rather than an exact ledger. If one channel consistently shows more store revenue per dollar than another under several models, that is a stronger signal than a single number in a single view.
Store revenue is the anchor. When the sum of platform-reported conversions is far above your real orders, the difference is mostly overlap and modelling. Decisions about total spend should be checked against what the store actually took in.
For bigger questions, such as whether a channel adds sales that would not have happened anyway, attribution alone is not enough. Holdout or geography tests, run on the platforms where they are offered, give a clearer answer and are worth considering once spend is significant.
Common attribution mistakes
Adding up platform totals
Each platform claims its own share of the same orders. The total is usually higher than real revenue.
Comparing different windows
A seven-day click window and a one-day click window give different numbers. Compare channels on the same settings.
Ignoring broken tracking
A theme update or a new checkout app can stop purchase events. Check events after every store change.
Treating one model as the truth
Last-click undervalues channels that introduce customers; spread models can overvalue them. Use more than one view.
Attribution data is partly modelled and never complete. Novaris does not guarantee the accuracy of platform-reported conversions or any result; outcomes depend on your products, market, budget, tracking setup and ad accounts.
Questions about attribution tracking
Why don't my platform numbers match my store?
Each platform counts conversions with its own model and window, and several platforms can claim the same order. Some conversions are also modelled estimates. Your store records each order once, which is why we use it as the reference.
Which attribution model should I use?
There is no single right answer. Last-click is simple and conservative, while multi-touch views give more credit to channels that start journeys. Comparing two views usually tells you more than picking one.
Can Novaris track every conversion?
No. Browser privacy features, ad blockers, consent choices and app tracking rules prevent some conversions from being tied to an ad. No tool can measure what was never collected.
What are modelled conversions?
They are estimates that ad platforms add to cover conversions they could not observe directly. They can be useful, but they are not counted orders, and we label them where the platform identifies them.
Is attribution available on Starter?
Starter includes basic analytics with revenue by channel and campaign. Growth and Scale add advanced analytics with deeper cross-channel views and more ways to compare models.
Do I need server-side tracking?
It is not required, but server-side events, such as the Meta Conversions API offered through many store apps, can recover some events that browsers block. We check what your store supports.
Does Novaris change how platforms optimize my ads?
No. Each platform optimizes delivery with its own data and models. Novaris reports on results and helps you decide where to spend; the platforms decide how they deliver.
Advanced analytics
Real-time cross-channel reporting to base decisions on.
Open →Integrations
Connecting your store, pixels and conversion events properly.
Open →Budget optimization
Using attributed revenue to move budget within limits you set.
Open →General information, last reviewed in November 2026. Features depend on your plan; results are not guaranteed and depend on your products, market, budget and ad accounts. Ad platforms set their own policies.
Tell us about your store.
Tell us about your store and the channels you advertise on. We set up your trial and show you the platform on a short call.