Choosing an agency is mostly a memory test. You sit through four pitches over three weeks, they all show a nice deck, they all say they understand Microsoft Partners, and by the time you're deciding you're comparing a feeling about a person against a slide you can half remember.
That's how partners end up twelve months in, paying a retainer, with a lot of activity and no answer to "what did this produce".
So do it differently. Score it. Same criteria for everyone, written down before the first call, weighted before you've met anyone you like. It takes an hour to set up and it removes most of the risk.
Here's the method we'd use if we were on your side of the table.
First, decide what you're actually buying
"We need a marketing agency" isn't a brief, and it's the reason so many of these processes go sideways. There are at least four different jobs hiding in that sentence, and the best agency for one is often a poor fit for another.
A thinking job
Your proposition is vague, you sound like every other partner in your solution area, and nobody internally can say in one sentence why a customer should pick you. What you need is strategy and value proposition development, not a content calendar.
A pipeline job
The proposition's fine, sales are fine, there just aren't enough conversations. That's demand generation, and it should be judged on pipeline, not impressions.
A capacity job
You know what to do, there's one marketing person doing it and they're drowning. You want scalable marketing support that plugs in and runs, not a strategy workshop.
A Microsoft job
You want to be visible to Microsoft sellers, get co-sell moving, and stop leaving co-op funding on the table. That's a specific skill set and most agencies don't have it.
Write down which one is 70% of the value. Score against that.
The scorecard
Twelve criteria, weighted. Score each agency 1 to 5, multiply by the weight, total it. Adjust the weights to match the job you defined above, but do it before the first meeting.
| Criterion | Weight | What a 5 looks like | What a 1 looks like |
|---|---|---|---|
| Microsoft fluency | High | They use co-sell, solution areas, specialisations, incentives and marketplace correctly and unprompted, and can explain what changed this fiscal year | Microsoft is described as "a vendor partnership" and the specifics stay vague |
| Proportion of Microsoft clients | High | Most of their client base is Microsoft Partners, and they can describe the differences between an ISV, an MSP and an SI without prompting | You'd be their Microsoft experiment |
| Evidence of outcomes | High | Named numbers tied to pipeline or revenue, with the context of what they did and how long it took | Award logos, follower growth, and "engagement is up" |
| Who does the work | High | You meet the day-to-day team in the pitch and they contribute | Impressive senior people who won't be seen again after signature |
| Funding literacy | Medium | They can scope a campaign so the claim and evidence are planned from the start, and know the deadlines that apply to you | "We can help you claim MDF" with no detail behind it |
| A point of view on your business | Medium | They arrive having found something specific about your positioning that's wrong, and say so | A generic capability deck with your logo on slide one |
| Measurement model | Medium | Clear on what they'll report, how attribution works, and what they'd expect at 90 days versus 9 months | Monthly activity reports with no line back to pipeline |
| Channel honesty | Medium | They tell you which channels won't work for you and why | Everything is recommended, because everything is billable |
| Content depth | Medium | Writers who can interview your technical people and produce something a CIO would finish | Volume-first content that reads like it was generated in bulk |
| Speed to first output | Low | Something real in market within the first six weeks | A three-month discovery phase before anything ships |
| Commercial flexibility | Low | Sensible notice, scope you can adjust, no penalty for changing direction | Twelve-month lock-in with a fixed scope |
| Cultural fit | Low | They challenge you politely in the pitch | Total agreement with everything you say |
Why "Microsoft fluency" carries the most weight
A generalist B2B agency can write well, run paid media and build a decent funnel. What they can't do quickly is understand the second customer you're selling to.
Because you've got two. There's the end customer, and there's Microsoft: sellers who could bring you into deals, account teams deciding which partner to trust in front of their customer, and a funding and incentives structure that quietly shapes what's worth doing this year.
An agency that only understands the first one will build you a perfectly competent programme that ignores your biggest source of pipeline. Getting that second audience right is most of what marketing into Microsoft involves, and it's the hardest thing to learn on your time and budget.
Test it directly. Ask what's changed in Microsoft's priorities this fiscal year and what it means for a partner like you. Vagueness there is disqualifying.
Ten questions that get you a straight answer
Ask all ten of every agency, in the same order, and write down the answers. The comparison does itself.
- What proportion of your clients are Microsoft Partners, and how many are in my solution area?
- Show me a piece of work that influenced a co-sell deal, and tell me what it took.
- Who exactly will be on my account every week, and what else are they working on?
- What would you do in the first 90 days, and what should I expect to see by day 90 that I don't see today?
- How would you scope a campaign so co-op or MDF funding is claimable, and what evidence would you collect as you go?
- What have you looked at on our website and in our positioning that you think is wrong?
- Which channel would you not recommend for us, and why?
- How do you report? Show me a real client report with the names removed.
- What's happened on a client programme that didn't work, and what did you change?
- What do you need from us for this to succeed, and what happens if we can't give you it?
The answers that should worry you
"We work across the whole tech sector."
Fine for a generalist brief. A problem when your pipeline depends on Microsoft sellers understanding your solution well enough to introduce you.
Results with no numbers
"We grew their brand awareness significantly" is not a result. If nothing can be shared, ask for one anonymised outcome with a real figure attached. Any agency doing serious work has one.
A pitch team that isn't the delivery team
Ask early, ask directly, and get the day-to-day names into the contract.
Everything is achievable
Good agencies tell you what won't work. An agency that agrees with all of it is selling, not advising.
No opinion on your proposition
If they've read your site and have nothing to say about it, you're buying execution capacity, which is fine if that's the job, and expensive if it isn't.
Funding treated as an afterthought
Retrofitting a claim after the campaign has run is where partners lose money. If they can't talk practically about scoping and evidence, they'll cost you more than their fee. The Microsoft Co-op Funds Hub sets out what qualifies.
Run a paid pilot instead of a beauty parade
The pitch process rewards presenting. It tells you almost nothing about the working relationship, which is what you're actually buying.
So shortlist two, then buy a small, paid, real piece of work from each. A proposition sprint. One campaign end to end. A content piece that needs a technical interview. Six weeks, defined output, sensible money.
You'll learn more in those six weeks than in any number of meetings: how they handle your subject matter experts, whether the work needs three rounds of rewrites, whether they chase you or you chase them, and whether the thinking is as good as the deck.
It costs less than nine months with the wrong agency, and it's the closest thing to a guarantee you'll get.
What to expect once you've signed
Set the expectation early, internally as well as with the agency, or you'll judge it at the wrong moment.
Weeks 1 to 6
Positioning sharpened, plan agreed, and something real in market. If nothing has shipped by week six, that's a signal.
Day 90
Two or three live plays, funded activity underway if it applies, reporting in place, and early signs of the right conversations. Not a full pipeline. B2B cycles in this space are long and anyone promising otherwise is guessing.
Months 6 to 9
Pipeline you can attribute, at least one reference customer produced by the work, and a proposition your sales team uses without being asked.
Agree those checkpoints in writing before you start, along with what you'd both do if the numbers aren't there. It's a much easier conversation to have in month one than in month seven.
The short version
Define the job before you take a single meeting. Score everyone on the same twelve criteria and weight Microsoft fluency heavily. Ask the same ten questions and write down the answers. Pilot the shortlist rather than picking from a deck. Agree what day 90 looks like before you sign.
Do that and you'll pick on evidence rather than on who presented best on a Thursday afternoon.
And if you'd rather start with the honest version of where your marketing is now, we'll tell you what we'd change before you commit to anything.