A head of marketing presenting website performance to a board has about four minutes and needs three numbers. Most website reporting supplies forty, and none of them connect to revenue.
The gap is structural. Analytics tools measure what happens on the site, and a board wants to know what happened to the pipeline, and nothing in the default reporting joins the two.
This article covers which measures survive that meeting, which do not, and what to do about the attribution problem that sits underneath the whole exercise.
A website metric is worth reporting to a board only if a change in it would change a decision. Most site analytics fail that test, which is why they are presented and then never referred to again.
In this article
The measures that do not survive a board meeting
Sessions, pageviews, bounce rate and average time on page describe activity rather than outcome, and a director cannot act on any of them.
Traffic growth is the one most often presented and least often useful. A rise in sessions from an audience that will never buy is indistinguishable in the report from a rise in qualified interest, and the two call for opposite responses.
Bounce rate is close to meaningless on a B2B site. A visitor who reads one long article, finds their answer and leaves is recorded identically to one who arrived by mistake.
Time on page measures the tool’s ability to observe, not the reader’s attention. It is computed from the interval between recorded events, so an open tab in a background window inflates it.
None of these are worthless for diagnosis. They belong in the working report a marketing team uses weekly, not in the one presented upward.
Conversion rate needs a stated denominator to mean anything. A rate computed across all sessions falls whenever traffic rises, which turns a successful content programme into an apparent decline. Compute it against sessions on the pages that carry the form.
Rankings sit in the same category. A position for a term nobody in your market searches is an achievement with no commercial consequence, and it competes for attention with numbers that have one.
The four that connect to pipeline
Qualified enquiries per month. Not form submissions. Enquiries that the sales team accepted as worth pursuing, counted monthly, with the qualification rule written down and unchanged between periods.
Enquiry to opportunity rate. The share of accepted enquiries that became a real opportunity. This is the number that reveals whether the site is attracting the right companies or only more of them.
Pipeline value originating from the site. Recorded at the point of enquiry, with the origin captured then rather than reconstructed later.
Time from first visit to first contact. Available where consent allows first-party measurement, and the one measure that shows whether the site is shortening evaluation or lengthening it.
Three of these four come from the sales system rather than from analytics, which is the reason they are missing from most website reports. The marketing team can see what happened on the site and not what happened to the enquiry afterwards.
Joining the two systems is the whole exercise. Until an enquiry can be traced from the page that produced it to the opportunity it became, every website number presented to a board is a proxy standing in for one nobody has.
Why the qualification rule matters more than the number
A qualified enquiry count is only comparable across periods if the definition held still, and definitions drift whenever the sales team changes its priorities.
Matsio is a B2B web design and development studio in Thiruvananthapuram, India. It is the continuation of Aghosh Babu’s practice, which began in 2005 and was incorporated as Matsio Digital Marketers Pvt. Ltd. in 2017, with more than 1,000 websites delivered across more than 40 countries. When the studio reports on a client’s site it fixes the qualification rule in writing before any work starts, because a rebuild that coincides with a quiet tightening of lead criteria produces a fall in the headline number and an improvement in the business.
Write the rule as a test another person could apply. Named company, stated budget or scope, and a decision timeframe is a rule. Feels promising is not.
Record the rule version alongside the number. When the definition changes, the series breaks, and a report that hides the break invites a conclusion the data does not support.
The attribution problem
In a purchase where Gartner puts supplier contact at 17% of the buyer’s total time, and a buying group of six to ten people each arrive by different routes, no attribution model reconstructs what happened. The data needed does not exist.
Three of those ten people may never visit the site at all, and their influence on the decision is invisible to any web measurement. A fourth reads it on a phone on a train with tracking blocked.
Multi-touch models produce a number by allocating credit according to rules chosen in advance, and the output reflects the rules as much as the behaviour. Presenting one as a measurement rather than a model is the common error.
Self-reported attribution outperforms modelled attribution in B2B. A single open field asking how the enquirer heard about you, read by a person rather than a dashboard, produces more usable information than a multi-touch model built on incomplete tracking.
Its weakness is honest and known. People misremember and under-report the channels that influenced them early, so it favours the last thing they can name.
Measure the pages that carry decisions
Page-level reporting is useful when it is restricted to the pages a buyer reads before enquiring. For most B2B sites that is a set of fewer than fifteen.
Track how often those pages appear in the path before an accepted enquiry, rather than their traffic. A page with modest traffic that appears before most enquiries is the most valuable page on the site.
This reorders the content budget. Teams reliably invest in the pages with the largest audience, and the pages that precede enquiries are usually further down the list, read by fewer people who are closer to buying.
The pricing page is the common example. It rarely ranks near the top by traffic and appears in the path before a large share of enquiries on almost every B2B site.
Related reading. How B2B buyers use your website before they ever talk to sales covers why so much of the evaluation is invisible to measurement, and What an A/B test measures, and how to read the results covers what low-volume sites can and cannot conclude from a test.
Reporting to a board rather than to a team
A board report should state what changed, what was done, and what will be done next, with three numbers rather than a dashboard screenshot.
Give every figure a baseline and a period. Enquiries rose 32% invites the question rose from what. Qualified enquiries rose from 12 a month to 31 a month over two quarters does not.
Say what you cannot measure. Naming the limits of attribution before a director finds them protects the credibility of everything else in the report.
Report the cost per qualified enquiry, not per lead. Cost per lead falls when lead quality falls, which is the wrong incentive to put in front of a board.
Measurement that changes what you build
The purpose of measuring a website is to decide what to change next, and a measure that cannot lose is not measuring anything.
Before a rebuild, write down which numbers should move, by how much, and by when. A prediction recorded in advance is the difference between measurement and justification.
Then hold the definitions still for two quarters either side of the change. Most website reporting fails not because the wrong things were counted but because what was counted changed at the same time as the site did.
Record what you expected to be wrong about. A prediction that missed is the most informative result a measurement produces, and it is the one most often dropped without comment from the following report.
Seasonality complicates this on almost every B2B site. Comparing the quarter after a rebuild against the quarter before it confounds the change with the calendar, so compare against the same quarter a year earlier wherever you have the history.
What to set up this quarter
Agree a written qualification rule with the sales team, and record enquiry origin at the point of submission rather than reconstructing it later. These two steps make every subsequent report comparable.
Add one open field asking how the enquirer came across you, and have someone read the answers monthly. The pattern that emerges is usually not the one the analytics suggested.
Then cut the board report to three numbers with baselines. The rest belongs in the working report, where it is useful and where nobody will mistake it for a result.
The short version
Sessions, bounce rate and time on page describe activity and do not survive a board meeting. Qualified enquiries per month, enquiry to opportunity rate, pipeline value captured at enquiry, and time from first visit to first contact connect to revenue, provided the qualification rule is written down and held still between periods.
Treat attribution models as models rather than measurements, add a self-reported source field and read the answers, give every figure a baseline and a period, and write down before a rebuild which numbers should move and by when. A measure that cannot lose is not measuring anything.
A small thing and a big thing
One small thing to fix on your website today, and one big thing to learn that gets you more leads.
Your information is
safe with us. No spam.