Your first deliverable in a new marketing job is a list of questions, not a deck. That's unpopular in week one, when the calendar already carries launch dates somebody committed to before you signed, and the fastest way to look useful is to produce something. It's also the fastest way to be wrong.
Gallup found that only 12% of employees strongly agree their organization does a great job onboarding new employees, and that turnover can run as high as 50% in the first 18 months! Most of us get no real handover. So the structure has to come from us, and it takes the shape of what we ask and who we ask it of.
Writing positioning before you understand the product is like drawing a map of a city you've only seen from the plane. The outline comes out roughly right, every street is wrong, and nobody finds out until somebody tries to use it to get somewhere. So what do you actually ask, in what order, and who owns the answers?
Why is week one the cheapest week you'll ever have?
Because understanding is the part of the work that everybody agrees with in principle and compresses in practice. It looks like preparation rather than output, so it's the first thing anyone trims when a launch date moves, and it's the one part where compressing it guarantees you pay later. The bill arrives around month six, when the messaging framework everybody signed off on turns out to describe a buyer who doesn't exist and never did.
In deep tech the bill is bigger. The buyer is often an engineer, an architect, or a data scientist, and they read your website the way they read a spec sheet. If the value claim is thin, they don't argue with it. They close the tab.
That's true whether you're the first marketing hire, a founding marketer, or a fractional consultant with a twelve-week scope. The fractional version is harder, because nobody is paying you to sit and read, so ask anyway, in the first days, while being new is still a license to ask anything.
Who do you ask, and what do you ask them?
Whoever decided there was a need
Usually a founder, an owner, or the VP of Product who wrote the job description, and a vague answer here is data in itself.
The product
You learn a product by using it, not by reading the deck somebody made about it. Ask for a license, a sandbox, and the same onboarding a customer gets, then run the workflow end to end and write down every place you got stuck. Those notes are the closest thing to a new buyer's experience anyone in the building will have for another year.
People already talking to buyers
Sales, customer success, and support already hold most of the answers that marketing spends its first month guessing at. Why five recorded calls and not one? Because one call is an anecdote, five is a pattern, and you can't hear a pattern in a summary somebody else wrote for you.
The people who inherit it
Product managers, engineers, technical writers, and designers all inherit your framework and any of them can ignore it without ever saying so, so ask them in week one and they own it with you. Read what is internal, too, not only what is published: the specs, the support tickets, the Slack thread where somebody explains the product properly to a colleague.
What does that look like on a calendar?
DayWhere the time goesWhat exists by the end1Kickoff with whoever opened the roleGoals and deadlines, in writing2The product, hands on, plus the internal docsStuck points and a glossary3Recorded buyer calls, sales, supportObjections, competitors, buyer words4Product, engineering, docs, designDifferentiators split from baseline5Writing it up, then reading it backA one-page brief and a list of unknowns
Day five is the one people skip, and it is the one that pays for the other four. Reading your understanding back to the people who gave it to you catches the misunderstanding while it is still free to fix, and it gets you the correction in front of witnesses.
What do you do with the answers?
They go into one document, not five slides. The positioning and messaging brief keeps the same spine every time, whether the team reading it is one founder or the 13 direct reports I had at Coro, and the questions above map onto it directly.
Does answering all of it guarantee that the positioning is right? Of course not. It guarantees the positioning is checkable, which is a smaller claim and a more useful one, because every line in it traces back to a person who said it and a call you can replay. Positioning built on a guess is positioning you rewrite in six months, and the same first-week sequence sits underneath the GTM framework work case.
Then act on what comes back, including the times when the answer is inconvenient. At Visual Layer, a computer vision company, we had a product-led growth motion underway, and after a product-market-fit analysis with the founders I cancelled it and moved us to account-based marketing and direct enterprise sales. ARR went from under $1M to $3M, and Camtek acquired the company in April 2026.
Years earlier at Amdocs, a manager told me he knew I could do the job and hadn't seen it yet. He was right. What he wanted was somebody who understood the value of the product before writing a word about it, and across 25 years in the industry that order of operations hasn't changed.
Key takeaways
If you are starting somewhere next month, or you have just hired someone who is, bring that first week to a free 30-minute call. We'll go through the list together, cut the questions your product does not need, and you'll leave with a version your team can hand to the next person who walks in.
