Unified marketing dashboards: what your board should be able to see

Your board needs four answers from marketing: what pipeline it produced, what a qualified lead costs by channel, whether the forecast still holds, and how confident you are in the numbers. Most marketing dashboards answer none of them, because they were built to report activity. The fix starts with agreed definitions rather than a choice of tool, and it ends with five views instead of forty.

What does a board actually want to see?

One question in three parts. Is marketing producing enough qualified demand to hit the number. What is it costing per unit. And how confident should we be in the measurement.

Everything else is a supporting detail that earns its place only when it answers one of those. Directors are not interested in channel performance for its own sake. They become extremely interested in it at the moment it explains why the top line moved, and at no other moment.

That is why the usual marketing dashboard reads badly in a boardroom. It was designed for the person running the channels, who needs to see everything, and then handed to a group of people who need to see four things and understand the fourth.

Why do most marketing dashboards fail the board test?

Because they report what happened and stop there.

A dashboard showing pipeline down 15% is a smoke alarm. It tells you there is a problem and nothing about which room. Somebody then spends three days in spreadsheets producing an answer that arrives after the decision has been taken, which teaches the board that marketing reporting raises questions rather than settling them.

The practical test is whether a reader can get from the number to the cause without leaving the view. Which channel, which segment, which week. If that takes a second dashboard, a data request or an analyst, the reporting is incomplete however well designed it looks.

Two other failures show up almost as often. Too many tiles, so nobody knows where to look and everyone looks at whichever one is red. And no named owner, so when a number is questioned there is nobody in the room whose job it was.

Which five views are worth building?

Five, and no more. Anything else belongs in the operational reporting your team uses on a Monday.

One, pipeline contribution. The value of open and won pipeline where marketing created the opportunity, shown separately from where marketing influenced it. Two numbers, two definitions, both written on the view itself along with the attribution rule being applied. This is the most argued over view in B2B reporting, and almost all of the argument disappears when created and influenced stop being blended into a single figure that flatters everybody.

Two, cost per qualified lead by channel. The unit economics view, and the one a finance director will read first. It needs a single definition of qualified, agreed with sales, applied to every channel including the ones that are inconvenient. Show it as a trend rather than a snapshot. The direction of that line over four quarters tells the board more than its level in any one of them.

Three, the signal ladder. The early readings that predict the later ones. Attention on the same day, intent at three to seven days, qualification at 14 to 30 days, revenue at 60 to 180. This is the view that lets you say something honest and useful about a programme five weeks in, rather than showing an empty revenue column and losing the argument. It comes straight out of Kill Perfect, the testing system I set out at Webflow Conf in Boston on 2 September, and it is the view most marketing dashboards are missing entirely.

Four, forecast against target. Where you expect to land, the assumptions underneath it, and how that forecast has moved since last quarter. Show the previous forecast alongside the current one. A forecast with a visible track record earns more trust than a confident one with no history, and a marketing team whose forecasts have been roughly right for three quarters is in a different negotiating position at budget time.

Five, one honest view of what cannot be measured. The activity you fund that produces no trackable click. Events, partner marketing, video, the conversation at a stand, the slide someone screenshots in a webinar. Professor John Dawes at the Ehrenberg Bass Institute, writing for the LinkedIn B2B Institute in 2021, put the share of B2B buyers in market at any given moment at around 5%, and most of the work aimed at the other 95% lands here. Give it a named budget line, state the proxy you use, and state your confidence in it.

That fifth view is the one people leave out, and it is the one that protects the rest. A board handles acknowledged uncertainty far better than it handles a number that later turns out to have been a guess with a decimal point on it.

What should come off the dashboard?

Start with anything that has never once changed a decision. Remove it and see who complains. Usually nobody does, and the ones who do are describing an operational need rather than a board need.

Specifically: impressions on their own, social follower counts, email open rate now that mail privacy features have made it unreliable as a comparison over time, website sessions with no segmentation, MQL volume with no quality reading beside it, and any percentage shown without its denominator. That last one causes more board level confusion than the rest combined, because a 40% improvement on eleven leads and a 40% improvement on eleven hundred are the same shape on a chart.

Also remove anything the team cannot explain the derivation of within thirty seconds. A number nobody can trace is a liability in a room full of directors, and it will be the one that gets questioned.

How often should it refresh, and who owns the number?

Refresh at the speed of the decision, not the speed of the data.

Media and site metrics can refresh daily, because somebody acts on them daily. Qualified lead and pipeline views refresh weekly, which matches the sales rhythm they are read against. The board view refreshes monthly and then freezes. A board view that has changed between the pack going out and the meeting starting destroys trust faster than a poor number does, because it makes every other figure feel provisional.

Ownership is the part most teams skip. Each of the five views needs one named person inside the business, not at the agency, whose job it is when the number is wrong. The test is simple. If pipeline contribution is questioned on Thursday, whose diary does that appear in on Friday. If the answer is nobody, that view is decoration.

The person who owns the number and the person who produces the report can be different people. The person who owns the number and the person who has to defend it at the board should not be.

What does one agreed set of definitions make possible?

HMS Networks sells industrial communication hardware into operational technology teams across seven markets. The reporting sits in one model on infrastructure HMS owns, which is why numbers from different systems can appear in the same sentence without anyone reaching for a caveat.

The programme produced 1,588 sales ready leads from the website and 201 confirmed leads in CRM, of which 157 sit with the HMS own brands and 44 with other divisions.

That split of 157 and 44 is the part worth pausing on. It exists because somebody decided in advance how a lead reaching a different division would be counted and who it belonged to. Without that decision, the same programme produces two defensible totals, a meeting about which is correct, and a board that quietly discounts both.

The rest follows from the same discipline. Cost per qualified lead fell from around £760 through trade events to around £72, and the programme returned 35:1 on media. Both are statements about units everybody had already agreed to define the same way, on measurement the client owns rather than reporting held inside the platforms buying the media. A dashboard cannot create that agreement. It can display it, or it can display its absence in high resolution.

What does a good board dashboard feel like in the meeting?

Short. The marketing item takes eight minutes rather than twenty five, because the pack answered the standing questions before anyone sat down.

The questions change too. They stop being about the data, which is where they go when definitions are unsettled, and start being about the market. Why is cost per qualified lead rising in one region and flat in another. Whether the forecast assumes the same conversion rate as last quarter. What the fifth view's confidence rating is based on this time. Those are the questions you want, and you only get them once nobody is arguing about whose number is right.

Where to start

Three things this quarter, none requiring a purchase.

Ask three people, one from marketing, one from sales and one from finance, to write down the definition of a qualified lead independently and without conferring. Compare the three answers in a room. The gap between them is your first project, and finding it cost you twenty minutes.

Take your last board pack and mark every number nobody has acted on in the past year. Delete them from the next one and see what is missed.

Name an owner for each remaining view, and put the review in that person's diary rather than in a shared calendar nobody reads.

Our unified marketing dashboards work builds the five views above on definitions agreed first, and the model underneath them is data unification on infrastructure you own.

Frequently asked questions

What should a marketing dashboard show a board? Five views. Pipeline contribution with created and influenced shown separately. Cost per qualified lead by channel, trended. The signal ladder of early readings that predict revenue. Forecast against target with the previous forecast alongside. And one honest view of the activity that cannot be tracked, with the proxy you use and your confidence in it stated.

What marketing metrics should a board not see? Impressions alone, follower counts, email open rate now that mail privacy features have made it unreliable over time, unsegmented website sessions, MQL volume without a quality reading beside it, and any percentage without its denominator. As a general rule, remove any metric that has never changed a decision, and any number whose derivation the team cannot explain in thirty seconds.

How often should a marketing dashboard refresh? Match the refresh to the decision. Media and site metrics daily, because somebody acts on them daily. Qualified lead and pipeline views weekly, matching the sales rhythm. The board view monthly, and then frozen. A board view that changes between the pack going out and the meeting starting damages trust more than a disappointing number does.

Who should own marketing reporting numbers? One named person inside the business for each view, not at the agency. The test is whether a question about that number on Thursday lands in somebody's diary on Friday. The person producing the report and the person owning the number can differ, but the owner and the person defending it at the board should be the same.

Why do marketing dashboards fail at board level? Because they report what happened without explaining why it happened. A dashboard showing pipeline down 15% tells a board there is a problem and nothing about which channel, segment or week caused it. If getting from the number to the cause needs a second dashboard or three days of analysis, the reporting is incomplete regardless of how it looks.