There are few things in business more reassuring than a graph going up and to the right. Unfortunately, the graph might be showing impressions. It could be showing reach, website sessions, video views or one of the approximately four thousand other numbers marketing platforms have become very good at producing, but the important bit is that it’s going up, somebody has made it green and everybody in the meeting can feel reasonably pleased with themselves.
The problem is that none of that necessarily tells you whether the marketing actually did anything. You can have more traffic, more engagement, more impressions and a beautifully improving click-through rate while generating absolutely nothing that matters to the business. The numbers aren’t necessarily wrong, they’re just answering a completely different question, and somewhere along the way we’ve become remarkably good at reporting on marketing activity while forgetting to report on what happened because of it.
Marketing reports are brilliant at proving marketing happened
Most marketing reports begin with activity because activity is easy to measure. Google Ads knows how many times an advert appeared and how many people clicked it, LinkedIn knows how many people saw a post, GA4 knows how many sessions arrived on a website and an email platform can tell you how many people opened something. Put all of that together, add a few graphs and percentages, and you’ve got something that certainly looks like a report.
None of those numbers is useless. In the right context they can tell you an enormous amount about what’s happening within a campaign, but somewhere along the way we’ve started treating the measurement as the outcome. An impression becomes evidence of awareness, a click becomes evidence of interest and a form submission becomes evidence that marketing is working, despite the fact that we haven’t actually established what happened afterwards.
If I tell you a campaign generated 438,295 impressions, you know that an advertising platform counted 438,295 impressions. You still don’t know whether the right people saw it, whether anybody remembered it, whether those people eventually visited the website or whether the campaign contributed a single pound to the business. That’s quite a significant amount of missing information for a number we’re apparently very excited about.
Everything is green. Excellent.
There is a particular species of marketing report where almost every number has a green arrow next to it. Traffic is up 12%, engagement is up 8%, impressions are up 23% and cost per click is down 6%, which sounds particularly impressive because money is involved. You can get halfway through one of these reports before realising nobody has mentioned sales.
This isn’t an argument against measuring marketing metrics. If we’re running paid advertising, we absolutely want to understand CPC, CTR, conversion rates and acquisition costs. If we’re working on search, rankings, visibility and organic traffic matter, and if we’re running social media, reach and engagement can help us understand whether the content is actually reaching and resonating with anybody.
The problem is when that’s where the conversation ends. A useful report should connect those numbers to something the business actually cares about, because otherwise we’re just providing increasingly sophisticated evidence that people interacted with some marketing. That’s useful information, but it’s a long way from understanding whether the marketing worked.
“We generated 186 leads.” Great. Were any of them actually any good?
This is one of the biggest gaps between marketing reporting and commercial reality. A dashboard records a form submission as a conversion and suddenly marketing has generated a lead, but the dashboard doesn’t necessarily know that the person entered asdf@asdf.com, wanted a job, was trying to sell you office furniture or had absolutely no intention of spending any money. As far as the marketing platform is concerned, job done.
The sales team knows differently. The CRM probably knows eventually, and the finance system definitely knows if that person becomes a customer, but the marketing report can remain blissfully unaware and continue celebrating its 186 conversions. That creates a fairly obvious problem because if we’re optimising marketing based on the number of leads rather than the quality and eventual value of those leads, we can become extremely efficient at generating more of something the business doesn’t particularly want.
A lead is an event, whereas a customer is an outcome, and there can be an enormous amount of useful information between the two. Understanding what happens in that gap changes the conversation from “how many leads did we get?” to “what kind of business did the marketing actually create?” One of those questions is considerably more useful when you’re deciding where the next chunk of marketing budget should go.
The platforms are all marking their own homework
There is another slight complication with modern marketing reporting. Every platform would quite like you to believe it is doing a brilliant job, and conveniently, every platform has its own collection of numbers with which to demonstrate precisely that. Meta has its attribution model, Google Ads has its attribution model, GA4 has another view of the journey, LinkedIn has its numbers, your email platform has its numbers and Search Console has another collection entirely.
Individually, each platform can be incredibly useful. The trouble starts when we expect any one of them to explain the whole customer journey, because customers have an irritating habit of not behaving like attribution diagrams. They don’t politely choose a marketing channel, interact with it once and then purchase in a way that makes the monthly report nice and tidy.
Someone might first encounter your business through a LinkedIn post, Google you three weeks later, read an Insight, disappear again, see a retargeting advert, return directly, look at a case study and finally submit an enquiry two days later. You can force an answer to the question of which channel generated that customer if you really want one, and plenty of attribution models will happily provide it. The more useful question is what role each part of that journey played in getting them there.
Marketing doesn’t happen in channels
Businesses organise marketing into channels because it makes everything easier to manage. There is an SEO budget, a paid media budget, an email platform, a social media plan and perhaps somebody responsible for the website. The customer doesn’t experience any of those things as separate departments, though, and certainly doesn’t care how you’ve divided the budget internally.
They experience a business. They might discover you through search, become familiar with you through social, return because of an advert, trust you because of a case study and finally enquire because the website made the next step obvious. Afterwards, email might keep the relationship moving until they’re ready to buy, while the CRM records what eventually happened.
Reporting each of those things independently can therefore produce a completely distorted picture. The channel that creates the initial demand might receive no credit for the eventual conversion, while the channel responsible for the final click looks like the hero. If you make budget decisions from that information alone, you can quite easily turn off the thing that was making everything else work.
A 42-page PDF isn’t insight
There is a temptation in reporting to confuse volume with sophistication. If the report contains enough graphs, tables, screenshots, percentages and acronyms, it must surely contain something useful, and if it runs to 42 pages it must contain an absolutely enormous amount of useful stuff. Then somebody spends an hour presenting it because nobody could reasonably work out what any of it meant on their own.
Good reporting should make a complicated marketing operation easier to understand. If somebody needs to narrate every chart to explain why it matters, there’s a reasonable chance the report is documenting data rather than communicating insight. The useful bit isn’t necessarily the graph, it’s understanding what happened, why we think it happened, whether it matters and what we’re going to do about it.
Traffic fell 14%, for example, but that number doesn’t tell you enough on its own. Did branded search drop, did rankings disappear, did seasonality change or did a campaign end? Did enquiries fall with it, did revenue change and was the traffic we lost commercially valuable in the first place? Without that context, the percentage is just a percentage with a red arrow next to it.
Reporting should occasionally contain bad news
A useful marketing report shouldn’t be designed to make the marketing team look good. It should be designed to make the next marketing decision better, and sometimes that means saying something isn’t working. That shouldn’t be controversial, although judging by some marketing reports you’d think admitting a campaign underperformed required a formal announcement to the entire company.
Paid search might be generating plenty of leads but hardly any of them become opportunities. Organic traffic might have grown substantially while most of that growth is coming from informational searches with little commercial value. Social reach could have doubled while website visits from social barely moved, or email engagement might look perfectly healthy while the database itself hasn’t grown for six months.
Those aren’t reporting failures. Finding them is the entire point of reporting, because once you understand what isn’t working you can actually do something about it. If every metric is positive every month, either you have accidentally assembled the greatest marketing operation in recorded history or somebody is choosing the metrics rather carefully.
The interesting bit happens after the conversion
Marketing reporting traditionally gets very excited about the moment somebody fills in a form. That’s understandable because it’s one of the easiest moments to track, but commercially it’s often the beginning of the interesting part rather than the end. A conversion tells us somebody did something, but it doesn’t necessarily tell us whether that something was worth anything.
What happened to that enquiry matters considerably more. Was it qualified, how quickly did somebody respond, did it become an opportunity, what was it worth and did the opportunity eventually close? If the person became a customer, did they spend £500 or £50,000, and did they disappear immediately afterwards or buy again six months later?
Suddenly the question isn’t simply whether Google Ads generated 40 conversions. It’s whether those 40 conversions produced customers and whether the value of those customers justified what we spent acquiring them. Ten expensive leads that produce four excellent customers can be considerably more valuable than 100 cheap leads that produce nothing except a very tired sales team.
Your CRM knows things your advertising dashboard doesn’t
This is why marketing data becomes much more useful when it stops living entirely inside marketing platforms. The CRM contains context that an advertising platform simply cannot have on its own because it knows what happened to the human being after they became a conversion. It can tell you who became an opportunity, what they were interested in, whether they progressed and, depending on how the business operates, whether they eventually became a customer.
Connect that information back to the marketing journey and suddenly the conversation becomes much more commercially useful. Instead of asking which campaign generated the most forms, you can start asking which campaign generated the most qualified opportunities. Instead of comparing channels purely on cost per lead, you can begin comparing what those leads were actually worth.
Those are rather different conversations, and they’re much closer to the conversation the person paying for the marketing probably wanted to have in the first place. Nobody has ever gone into business because they had a lifelong ambition to improve their click-through rate by 0.7%. They wanted more customers, more revenue, more profitable growth or some other commercially meaningful result.
Attribution will never be perfect
There is a slightly uncomfortable truth here that marketing technology occasionally tries to hide. You’re probably never going to know exactly what caused every customer to buy, because people talk to colleagues, see things without clicking them, switch devices, reject cookies and remember brands from something they saw months ago. Human beings remain stubbornly difficult to squeeze into a dashboard.
There is no magical attribution model that can reconstruct every one of those moments with absolute certainty. That doesn’t mean attribution is pointless, but it does mean we should stop pretending precision and accuracy are the same thing. A report can contain an impressively precise number based on a very incomplete understanding of what actually happened.
The objective therefore isn’t perfect attribution. It’s enough reliable information to make better decisions, combined with enough common sense to understand the gaps. If the data tells us something with confidence, that’s useful, and if it only gives us part of the picture, the report should be grown-up enough to say so.
The report should answer “so what?”
This is probably the simplest test for whether a marketing report is actually useful. Every important number should survive somebody asking “so what?”, because a metric without context doesn’t tell you whether you’re looking at an opportunity, a problem or something that doesn’t matter very much at all.
Organic traffic increased 22%, so what happened because of it? Paid conversions fell by 11%, so did sales fall too, and LinkedIn reach doubled, so did that create any meaningful increase in branded search, website traffic or enquiries? Email click-through rate improved, but did anybody actually take the action we wanted them to take afterwards?
Sometimes the answer is that a metric is an early indicator and we need more time. Sometimes it tells us something needs changing, sometimes it confirms a strategy is beginning to work and sometimes the answer is that the number doesn’t actually matter as much as we thought it did. Reporting becomes useful when it helps you distinguish between those things rather than simply presenting all of them with equal importance.
Stop reporting on marketing. Start reporting on what marketing did.
The shift is actually quite small conceptually, but it changes the purpose of the report completely. Stop treating the marketing activity as the final thing being measured and start following what happened because of it. That means connecting visibility to traffic, traffic to behaviour, behaviour to enquiries, enquiries to opportunities and opportunities to customers wherever the data allows it.
It also means looking across channels rather than allowing each platform to present its own little version of reality. Search, paid media, social, email, the website and the CRM are not separate customer journeys simply because they happen to have separate logins. If they’re contributing to the same commercial objective, the useful reporting is the reporting that helps you understand how they’re working together.
That’s ultimately the thinking behind Pulse. We didn’t build it because the world desperately needed another dashboard full of colourful graphs, and we’re not particularly interested in giving businesses another place to watch numbers move around. The point is to bring the useful parts of the marketing picture together, connect them with what happens further down the customer journey and make it easier to understand what deserves more investment, what needs fixing and what probably needs stopping altogether.
Reporting shouldn’t exist simply to prove that the marketing team was busy. It should tell you what the marketing did, what happened next and what you’re going to do differently because of it. If all you’ve got at the end of the month is a graph going up and to the right, it would probably be sensible to check what the graph is actually measuring.
