Most of the marketing teams we work with have access to more data than they can process. Google Analytics, CRM, campaign data, advertising platforms, social media tools. The problem is rarely a lack of data.

The problem is not knowing what to ask.

There are three questions that, when a team cannot answer them with data, reveal a structural problem in their marketing analytics. All three are simple. Most companies cannot answer them well.

Question 1: Which channel generates customers, not just leads?

Most teams measure attribution at the capture point: which channel generated the completed form, the call, the registration. That data is useful but incomplete.

The channel that generates the most leads is not always the one that generates the most customers. And the channel that generates the most customers is not always the one with the best ROI once acquisition cost and customer lifetime value are accounted for.

A well-configured marketing analytics system connects campaign data to CRM and sales data. The goal is not a more complete dashboard, but the ability to answer: "If I invest an additional £10,000 in channel X, how many additional customers do I generate, at what cost, and with what average contract value?"

If your team cannot answer that question, it has an attribution problem. And that problem has a cost: in the diagnostic projects we carry out, between 25% and 40% of the marketing budget is typically reaching audiences that are already customers, duplicating messages the user already received, or investing in channels whose real contribution to the business was never properly measured.

Question 2: How long does it take a customer to buy from when they first discover you?

The purchase cycle is one of the most important pieces of data for any marketing strategy. It is also one of the most overlooked.

Knowing that 30% of your customers converted within 48 hours of first contact and 60% took more than 30 days has direct implications for how you design campaigns, what messages to use at each stage of the funnel, and how to correctly attribute conversions that arrive weeks after the first interaction.

Without this information, teams systematically overvalue direct-response channels — which capture the end of the cycle — and undervalue demand-building channels — which initiate it. The result is a chronic reallocation of budget towards the urgent at the expense of the effective.

To calculate the purchase cycle, you need to cross first-interaction data with final conversion data. In most companies, that cross-referencing requires integrating sources that were never designed to communicate: the CRM with the advertising platform, the web form with the email history, the web analytics with sales team data.

Question 3: What characteristics do your best customers have before they become customers?

This is the most advanced of the three questions and also the most valuable.

The best customers — those with the highest lifetime value, lowest churn rate, greatest propensity to refer new customers — typically display identifiable behaviours prior to conversion. Pages they visited, content they consumed, actions they took, characteristic interaction sequences.

If you can identify those patterns, you can design campaigns that attract more similar profiles. And you can focus the budget on the type of demand that genuinely matters to the business, not merely on conversion volume.

This analysis requires behavioural data connected to post-sale data. In practice, it means marketing and sales sharing information systematically — something that in many organisations does not happen for structural reasons rather than technical ones.

The recurring pattern

In the analytics diagnostic projects we carry out, we find a consistent pattern: companies have the data needed to answer all three questions. What is missing is the integration between systems and the willingness to ask the uncomfortable questions.

The first step is not technological. It is defining which marketing decisions are made on a recurring basis and what data would be needed to make them better. From there, the analytics architecture is built with purpose — not as an infrastructure project without an internal client.

Companies that manage to answer these three questions do not merely improve their marketing metrics. They change the nature of their internal conversations: from "we believe channel X is working" to "we know channel X generates Y type of customer, at Z cost, with W lifetime value".

That difference is what separates assumption-based marketing from data-driven marketing.