Your server-side tracking is fixed, but what about your reporting?

Once your server-side tracking is up and running, your tags are in order and your data is more complete, it is tempting to think the hard work is done.

And fixing your data collection is a big step forward. But in a lot of the businesses we work with, there is still a gap between having better data and being able to answer the questions the business actually cares about.

The finance lead is still building channel performance in a spreadsheet every month. Marketing is still pulling numbers from different platforms. And the board is still asking questions that cannot be answered from any one of the eight or nine places where the data lives.

That is not a problem with server-side tracking. It has done its job. It is just that better data collection is one part of the bigger picture.

Here are a few things to think about once your server-side tracking is in place.

1) Know whether you have a collection or attribution problem – or both

You might find that the revenue in your analytics does not match the backend, or that too much revenue is being attributed to Direct. Both are signs that something is not quite right, but they can have very different causes.

Collection loss is data that never arrived. Events are not being recorded properly, so the sale never makes it into your analytics. This is exactly the kind of problem server-side tracking is designed to address. By giving you more control over how data is collected and sent, you can make your measurement much more reliable.

Attribution loss is different. The sale has been recorded, but the source has not. The data is there, but you do not know where the customer came from.

Both are worth investigating, because fixing one does not necessarily fix the other.

On a recent audit with a new client, we found both happening at the same time. There was a gap between analytics revenue and backend orders, but around a fifth of revenue was also sitting in an unattributed Direct bucket.

Improving the tracking closed most of the first gap, and better UTM tagging and referral exclusions dealt with most of the second. But getting to one data source everyone trusts takes more than that. You need to model the backend and analytics data together, usually by matching transaction IDs, and that isn't easy to do at scale.

Before assuming everything comes down to tracking, look at what is happening to your attribution too. A surprisingly large amount of Direct traffic can come down to something as simple as a checkout flow recording itself as a referral source.

The important thing is to understand which problem you have before deciding how to fix it.

2) Reconcile your data before you start fixing it

One of the simplest checks you can run is to compare your analytics with your backend.

Look at revenue and order count for the same period and the same timezone. Then do the same by channel.

The channel-level view matters. A 15% overcount in one channel and a 15% undercount in another can cancel each other out in the total and make the overall number look fine.

Once you know there is a gap, work out what is causing it.

How much is revenue that was never recorded? How much is revenue that was recorded but not attributed correctly? And how much is simply a legitimate difference between the systems?

Before you start digging into the data, rule out the obvious things:

• Refunds, where the backend deducts them but analytics does not
• Different treatment of tax, shipping and discounts
• Order types that never go through the storefront, such as POS, phone and subscription orders
• Timezone differences that put the same order on different days in each system

Getting this right gives you a much clearer picture of where your tracking is working, where it needs attention and where the problem actually sits elsewhere.

3) The next step is bringing the right data together

Once you have reliable data coming in, the next question is what else you need to connect to it.

When a business cannot answer a question, the first instinct is often to look for better analytics. More often, there is simply another data source missing.

Take margin by category. Businesses ask for it all the time, but product cost data rarely sits in a marketing platform. It might be in finance, the ecommerce backend or a spreadsheet that one person maintains.

No reporting tool can give you a proper margin view without that data.

Bring the right source into the model and suddenly the question becomes much easier to answer.

The same applies to channel-level acquisition cost, cohort lifetime value, returns and plenty of other commercial questions. These are not necessarily missing features. They are questions that become answerable when you have the right data in the same place.

So when something cannot be answered, ask “what data are we missing?” before asking “what tool do we need?”

There is also a point here about how you bring those sources together.

From working with businesses handling sensitive data, we have learned that it is better to be specific about what you actually need rather than pulling everything in just in case.

On one project, that meant an explicit allow-list of 18 database tables out of 60, with IDs hashed before anything moved, and regional reporting based on postcode groupings large enough that individuals could not be identified.

It takes a little more thought upfront, but it means you know exactly what data you are using and why.

4) Validate the numbers before you roll it out

Better data is only useful if people trust it.  If someone opens a new dashboard, checks the first number they care about and finds it is wrong, you can lose trust very quickly.

Finance is particularly important here. They may not tell you the dashboard is wrong. They will just carry on using the spreadsheet they already trust.

Finance should be able to reconcile it with their figures. Marketing should understand how it compares with their platform reporting. Operations should recognise the numbers they use day to day.

If something does not match, work out why before rollout rather than in the meeting afterwards.  It might add a few weeks to the project but it can make the difference between a reporting platform people actually use and one that gets quietly ignored.

5) Be realistic about the timeline

Getting from fragmented data sources to a model the business can actually interrogate is not a two-week job.  There is the tracking audit, the data model, the additional sources, the reconciliation and the validation.

That is not a criticism of any particular technology. It is simply the nature of getting different systems to work together properly.  The projects that tend to struggle are the ones that expect a complete picture almost immediately.

The important thing is to understand that “our tracking works now” and “we can answer the board's questions” are not quite the same milestone.

Server-side tracking gives you a much stronger foundation. The next step is making sure you can actually use that data alongside the other information the business needs.

So what next?

Server-side tracking is an important part of getting your data into better shape. It gives you more control over collection and can close some of the gaps that make your reporting unreliable.

But once the tracking is working, there is another question to answer: what can you actually do with the data?

Start by reconciling your analytics against the backend. Understand whether any gaps are caused by missing data, attribution or legitimate differences between systems.

Then look at what other sources you need to answer the commercial questions the business is asking.  Finally, validate the numbers with the people who are going to use them before you roll everything out.

Because better tracking is not the end goal. It is the foundation for better reporting, better analysis and, ultimately, better decisions.

‍…And if your numbers still don't add up

Bringing ecommerce, marketing and financial data together is a big part of what we do at ASK BOSCO®.  And for businesses where the numbers still don't reconcile after the tracking work is done, ASK BOSCO® also offers a bespoke, specialist-led service. Our data and tracking specialists connect your marketing, ecommerce and finance data, reconcile it against your backend, and validate the numbers with your finance and marketing teams before anyone logs in. From there, we build the reporting the business is actually asking for, such as acquisition cost by channel, cohort lifetime value and margin by category, and stay involved as your needs grow.

If you’d like to find out more, please get in touch with one of our specialists at team@askbosco.io