Why deep tech marketing can't show the board what it's doing
Gold star (not deserved)
Most deep tech companies reach this point eventually, and it's rarely comfortable. Marketing has been running for a while. The budget has gone on campaigns, events and content. Then someone at a board meeting asks what all of it has produced, and the answer either doesn't exist or doesn't connect to anything the board cares about.
In most cases the marketing team is capable. The trouble is how the work gets measured. The attribution tools most teams use were designed for much shorter sales cycles, and deep tech stretches them well past what they can handle.
The standard attribution model assumes a short path: someone sees an ad, clicks through, fills in a form, goes into a nurture sequence and eventually buys. That works for sales cycles measured in days or weeks, and enterprise SaaS stretched it to months. In deep tech, an OEM design win routinely takes eighteen months to two years. The buying committee includes engineering, procurement, finance and the CTO's office. A single deal can involve a dozen interactions across half that committee before anyone fills in a form or answers an email. When the commitment finally comes, as a design win, an IP license or a long-term supply agreement, a standard CRM attribution model has no way to count it.
So marketing that is driving commercial results often can't prove it. A principal engineer might read a whitepaper six months before procurement talks begin, or a technical director might first see a demo at an event. A content programme can keep an OEM interested through an eighteen-month evaluation. Last-touch attribution records none of it.
I've run into this myself. We ran a long ABM programme aimed at Samsung when it was considering its own Exynos CPU cores, and our goal was to persuade Samsung that Arm was the better commercial bet. Arm CPUs still sit at the heart of Galaxy phones today.
I'll take the credit, although I don't know if I deserve it. The decision came years after our activity, through a process I was never fully inside, and plenty of the factors behind it had nothing to do with anything we put in market. When the cycle runs that long, I find it hard to separate cause from coincidence, and I ran the programme.
So the board sees spend and a pile of leads that don't seem to convert, while the marketing team watches accounts move forward with no way of showing it in the numbers.
We worked through this inside one of the largest marketing programmes in deep tech. At its peak it covered hundreds of OEM accounts across several product lines, co-marketing with ecosystem partners, developer communities and hyperscalers, and sales cycles where top-of-funnel activity and the revenue it led to could fall in different financial years. To get attribution working, we had to change what we measured and when we measured it.
We mapped influence across each whole account. Our headline metric became account engagement depth: how far into a target account our content, events and programmes had reached, and whether that reach was producing the early signals that predicted progress towards a design win. We also split the pipeline into tiers, separating early-stage influence from late-stage intent, so the board could see the long-term investment and the near-term return together.
With that in place, we could go into a board meeting and show what marketing had influenced as well as what it had spent: at which stage, in which accounts, and how that mapped to the pipeline the business was trying to close. The same data showed us which programmes were moving accounts forward and which were producing activity that never converted, so we knew where to put the budget.
The companies I've seen fix this built their measurement around how their own sales cycle works, instead of adapting a consumer or SaaS model. That costs more up front than buying another marketing tool. It also needs someone who understands both the technical side of measurement and what the board wants to know, and most deep tech marketing teams I've worked with don't have that person.
Dale Kaszycki is a co-founder of Latent