01
An MQL doesn't exist for sales
A rep doesn't live off a lead score of 75. They live off invoices. An MQL is an internal marketing metric, often optimised for its own number and detached from the pipeline. Chris Walker (Passetto, formerly Refine Labs) has spent years pointing out that the traditional MQL→SQL→Win funnel is metric theatre - most MQLs never become an opportunity, while marketing keeps reporting to sales with numbers sales doesn't want to see.
What to measure instead? Pipeline-sourced revenue - revenue from opportunities a given channel originated. Opportunity-sourced revenue - revenue from closed opportunities. Deal velocity - how long it takes to move an opportunity through the pipeline. Those are numbers a rep looks at and they mean something.
In practice: instead of a weekly '47 MQLs generated' report - a monthly 'the LinkedIn campaign put 8 opportunities into the CRM, 3 are in qualification, total pipeline value 240k'. Marketing stops showing activity and starts showing impact.
02
95% of your potential buyers aren't buying today
The Ehrenberg-Bass Institute (Byron Sharp, John Dawes), one of the most serious research teams in marketing, has published data for years showing that at any moment only about 5% of a B2B market is in an active buying process. The other 95% are future buyers - people who will buy in 6, 12, 24 months. Most B2B firms still spend their budget on that 5% (paid search, lead gen, demo requests) and leave the 95% untouched.
The remaining 95% is the territory of brand building: thought leadership, industry reports, a podcast, events, a presence at trade shows, distributing knowledge in the media. The KPIs here are share of voice, brand search volume, mental availability. It works slower, but without it your campaign aimed at the 5% lands in a void - nobody remembers you when the buying decision finally starts to happen.
In practice: consider the 60/40 split between brand and activation that Peter Field and Les Binet derived from IPA data. In most Polish B2B firms the ratio is reversed, 90/10 or worse. Checking your own is a good first step - it's often a genuine shock.
03
Last-click attribution in B2B is a bedtime story for adults
A B2B buying decision takes 6-18 months on average and touches a dozen-odd points of contact - from a LinkedIn post, through an industry report, a colleague's recommendation in a private DM, a mention on a podcast, all the way to the demo form on your site. Gartner notes that enterprise decisions involve a dozen-odd people on average, each with their own path to the information.
That final click on 'demo request' is a formality. The real work happened weeks earlier, in dark social - marketing activity invisible to any GA4 or attribution model. A conversation about your company in a private Slack. A reshare of your report on LinkedIn. A recommendation from someone who once read your work.
In practice: rather than playing in the sandbox of attribution models - run self-reported attribution in your contact forms. A simple 'How did you hear about us?' is often more accurate than any algorithm. Plus: track brand search as a separate KPI in Search Console. Its growth says more about marketing's impact than any funnel dashboard.
04
Your 'demo request' is a lagging indicator, not a channel
A demo request is the result of marketing activity, not marketing itself. If you measure only the number of demos a month, you see only the last step of a long road. It's like weighing yourself once a month and being surprised you don't know why it's going up or down.
Demo requests are really generated by things you don't measure directly: the accuracy of your positioning (does a visitor understand in 30 seconds what you do and who you help), brand strength (have they heard of you before, or are they discovering you from scratch), content quality (does the article the ad points to solve a real problem or just stuff a keyword).
In practice: besides supply metrics (how many demos we caught this month) measure demand metrics - what drives interest. Growth in organic brand search, newsletter retention, ICP decision-makers following your LinkedIn page, time spent on substantive articles. The first predicts the second 2-4 months ahead.
05
A buyer persona from 2015 is killing your sales
Nobody buys because they're a 'Senior Marketing Manager, 38, likes sport and jazzy weekends'. People buy because they have a specific problem to solve. Tony Ulwick (Outcome-Driven Innovation, the Jobs-to-be-Done concept) proposed a radical shift: instead of asking 'who buys', ask 'what job does the buyer want done'. That question changes everything - from positioning, through channel choice, to the language in your copy.
Nobody hires the maker of IKEA furniture screws to 'have screws' - they hire them to 'assemble the shelf on a Saturday morning without calling a fitter'. The same goes for every B2B category. A CMO doesn't buy 'a marketing automation platform' - they buy 'the peace of mind that the campaign won't fall over at 6pm on Friday when the board asks about results'.
In practice: drop the word 'persona' from your briefs, bring in 'buying context' and 'jobs-to-be-done'. Audit your copy: do we talk about what the product does, or about the job it helps the customer get done? The fastest CRM - weak. A CRM that doesn't force the rep to sit on Friday evening updating statuses after a day of meetings - much better.
06
AI won't fix your marketing if the process is broken
Generating 100 leads a day with AI outbound (Apollo, Clay, Smartlead, Instantly) while the CRM is leaky, lead scoring doesn't exist, the rep comes back from leave after three weeks and the phone number on the site goes unanswered out of hours - only scales the chaos. AI is a force multiplier, not a generator of sense. In a broken process, AI multiplies the breakage.
Every 'AI lead gen' pitch I see has one blind spot: it assumes a working sales process on the receiving end, ready to handle ten times more enquiries. In practice maybe 5% of firms are. The rest first need an audit and a process fix, and only then scaling.
In practice: the order is sacred. First a process audit (how a lead enters the firm, who, when, whether anyone responds), then a tooling audit (whether CRM, marketing automation and analytics talk to each other), and only then AI. Deploying AI out of order is a recipe for a burnt budget and a frustrated sales team.
07
Your media budget isn't your biggest problem
Most B2B firms I talk to complain the budget is too small. After looking at the first numbers it turns out the problem isn't traffic, it's the conversion at every step of the funnel that nobody measures.
The site gets 5,000 visits a month. Of those, 50 fill in the form. Of those, 12 make sense. Of those, 5 answer the phone. Of those, 1 becomes a customer. End-to-end conversion: 0.02%. If you lift the conversion of steps 3-7 by 20%, you don't need more traffic. You get twice as many customers on the same budget.
In practice: audit the funnel before you raise the budget. Each stage with its own conversion number. First to fix: the bottlenecks with the highest impact on the final bill. Often it isn't the campaign creative, but, say, the rep's response time to the first enquiry - 5 minutes gives a fourfold higher close rate than 30 minutes.