We have an internal document at Conversion Factory that we never share with clients. Zach built it after the third or fourth time we got blamed for results that we knew were caused by the founder ignoring our recommendations. It's a scoring rubric — five criteria, three tiers each — that we run every prospect through before we agree to work together.
This post is most of that document, made public. Five of the criteria. The two we're keeping under lock and key.
If you're a SaaS founder thinking about hiring a marketing or design agency, this is the most useful read you'll get this quarter — because almost no agency will tell you why they actually declined to work with you. Most just say "we don't think we're the right fit" and move on. We're going to show you the framework behind that sentence.
If you're an agency or service provider reading this, you'll recognize most of it. The interesting thing isn't the categories. It's how they cluster. Founders who score poorly on one usually score poorly on three or four — these aren't independent variables, they're symptoms of the same underlying thing.
Here's the framework.
Criterion 1: Impact — how open are you to drastic changes?
The question we ask on most discovery calls now: how open are you to drastic changes?
It seems too obvious to ask. Of course they're open. They reached out to us because something isn't working. But the answer is often "we're open to changes as long as they don't disturb anything we've already built." That's not openness. That's optimization on the margin.
We score impact in three tiers:
- High impact (great fit). Founder is in for a real shift. Repositioning the company, rebuilding the brand, restructuring the website from the ground up. They came to us because they know the current foundation is the problem, and they're ready to tear it down to fix it.
- Medium impact (workable). Founder wants meaningful improvement but not a full rebuild. We can lift the quality across the board, but we're working inside their existing positioning, brand, and structural constraints. Sometimes this is the right call. Sometimes it's a tell.
- Low impact (red flag). Founder wants us to make small surface-level changes — copy tweaks, color palette polish, a few new landing pages — while expecting outsized results. There's nothing wrong with small changes. There's everything wrong with the expectation that small changes produce large outcomes.
The unforgettable example: a founder once told us they wanted to 10x their revenue. We mapped out everything that would need to change to make that realistic — positioning, pricing page, conversion funnel rebuild, content engine, demo flow. They agreed in the kickoff call. Then once the engagement started, the first thing they asked us to work on was activation emails. Activation emails are fine. They are not going to 10x your revenue. The math doesn't work and we knew it from day one.
This is the most common founder mistake, and it produces the most disappointing engagements. If the impact you want is large, your willingness to change has to match. Wanting big results from small effort isn't a strategy — it's a wish.
The fix: Before you hire any agency, write down what you actually want to change about your company. Not what tasks you want done. What outcomes you're optimizing for and what you're willing to put on the table to get them. If your list of "what we'll change" is short and your list of "what we want to result" is long, the engagement will not work. Reset one or the other.
Criterion 2: Async workflow — can you write things down?
This one looks like a workflow preference. It isn't. It's the single most predictive criterion in the entire rubric.
We're an async agency. We default to written communication, batched updates, and scheduled deep-work blocks. We do calls when they're necessary, not when they're convenient. The reason we structure it this way isn't ideology — it's that the math doesn't work otherwise. If we took two 30-minute calls per week with every client we had, literally half of our time would go to meetings and we'd produce half as much work.
But here's the deeper thing. After running this model for years, we noticed something: founders who can't tolerate async workflow almost always fail on the other four criteria too. It's a canary.
Here's why. The skill that async demands — sitting with a problem long enough to articulate it in writing — is the same skill that produces good strategy. Founders who can write down "here's what I'm seeing, here's what I think the problem is, here are three questions I have about it" already know what they want. Their thinking is structured. Their feedback is precise. Their goals are well-defined.
Founders who can't do that — who need to "hop on a quick call" to figure out what they think — usually don't have clear thinking yet. They're using the call to do their own strategy work in real time, while paying for someone else's time as a thinking aid. The output of those calls is almost always a half-baked decision that gets reversed two weeks later, because the underlying thinking still wasn't done.
The story Zach tells about this: he had a client who emailed a question. He answered it thoroughly in writing. The client responded "we need to have a call." On the call, the client asked the same question. Zach pointed at the written answer he'd already sent and read it back verbatim. The client said "okay" and the call ended.
That was the last client call Zach took for a while. It crystallized something he'd suspected: founders who insist on synchronous communication aren't just preferring meetings — they're often unable to do the cognitive work of reading and processing written material. And if they can't process written material, they can't articulate their own thinking either, which means everything downstream of them — strategy, brief writing, feedback, prioritization — is going to be muddled.
The tiers we use:
- Async-native (great fit). Founder writes clearly, batches feedback, doesn't ping us during deep work hours. Communicates in Notion docs, Loom videos, or thoughtful Slack threads. Calls happen weekly or biweekly, not daily.
- Async-tolerant (workable). Founder prefers calls but can work async if we set the expectation. Some friction, but it improves over the first month.
- Async-resistant (red flag). Founder needs constant calls, can't tolerate written feedback loops, doesn't read documents we send, and treats response time as a measure of effort. This founder will not be happy with any agency.
The fix: Before your next engagement, try writing a one-page brief that articulates (a) what you're trying to accomplish, (b) what you've tried that didn't work, (c) what you specifically need help with, and (d) three questions you have for the agency. If you can do that in 45 minutes or less, you'll work well with any good async-default team. If you can't get past the first paragraph, that's diagnostic — and the diagnosis isn't about workflow preference, it's about strategic clarity.
Criterion 3: Timelines — innovation vs efficiency
We get asked about timelines on every sales call. Sometimes it's a reasonable question. Often it's the wrong question.
Here's the framing we use internally. Creative work falls into two categories that almost never overlap:
Efficient work. Predictable, scoped, low-creativity-required output. A demo page following an established pattern. A standard pricing page. A feature comparison table. We can quote these accurately because we've done a hundred of them. They can be done fast and they can be done well, because the path is known.
Innovative work. Brand identity exploration. Positioning work. Naming. Category creation. Anything that requires creative mess and divergent exploration. This work does not have a predictable timeline because the process is fundamentally iterative. You don't know what the right answer is until you've explored several wrong answers. Compressing the timeline doesn't make this work faster — it makes it worse.
The mistake founders make is asking efficient-work timeline questions about innovative work. "Can we have a complete brand system in three weeks?" Sure — you can have a complete brand system in three weeks. It just won't be a good one. You'll get the first version of the work that didn't get time to marinate, didn't get time to be argued with, didn't get time to evolve. You'll launch it, regret it in six months, and pay to redo it.
The other mistake: treating arbitrary deadlines as input rather than output. We've had founders ask us to commit to "site live by the end of the quarter." Why end of the quarter? Usually no real reason — a date pulled from a planning meeting that someone wanted to feel ambitious. We're the ones doing the work. We know what the work takes. The deadline should be our output, not our input.
Parkinson's law applies here brutally. Work expands to fill the time available. Give yourself a year to build a brand, you'll spend a year building a brand. Give yourself a month, you'll build something in a month — but the something will be worse, and probably worse than what you would have built in six weeks with breathing room.
The Basecamp guys have a take on this that we agree with: plans are guesses. A 12-month product roadmap is a guess. A six-week project timeline is a guess. The point of the guess is to orient effort, not to commit to a precise endpoint that turns into the wrong thing six weeks later.
What we look for instead: founders who treat timelines as estimates, give us transparency into the progress, and trust the work when it's taking the time it should take. The transparency is the substitute for the deadline. If you can see the progress bar moving — not "25%" but "25.57%" — you don't need a delivery date, because you know it's on track.
The tier breakdown:
- Estimates-and-trust (great fit). Founder asks rough timeline questions to plan their side, trusts our progress visibility, and doesn't pressure for arbitrary speed.
- Reasonable urgency (workable). Founder has a real external deadline (funding milestone, product launch, conference) and can articulate why the date matters. We can usually work backwards from that.
- Arbitrary-deadline-driven (red flag). Founder picks dates that have no real grounding, then pressures us to commit to them. Almost always combined with criterion 1 (low-impact requests with high-impact expectations).
The fix: When you ask an agency about timelines, lead with why the date matters, not when. If you can't articulate a real reason for the date, the date isn't useful. Replace it with milestones tied to outcomes and let the agency tell you what reasonable looks like.
Criterion 4: Expected results — silver bullets vs gold BBs
This is the criterion that produces the most spectacular fails. Founders come to us with revenue targets that bear no relationship to the budget they've allocated, the changes they're willing to make, or the timelines they expect. We've literally had this exact conversation more than once: "We want to 10x ARR in the next year. We can spend $50K on marketing to make it happen."
The math doesn't work. Not in any version of reality. And if you can't see why the math doesn't work, you're not ready to hire a marketing or design agency yet — you're ready to spend a quarter learning unit economics first.
Here's the metaphor we use. Most founders show up looking for a silver bullet. One thing — a campaign, a redesign, a paid channel — that will fix everything. The thing they don't realize is that silver bullets, if they exist at all, take seven months to forge and you get one shot. If it misses, you've burned seven months.
What actually works is a hundred gold BBs. Smaller bets. Each one is decent craftsmanship — not garbage, but not optimized to perfection. You fire each one and watch where it lands. You adjust your aim based on what you saw. Some hit, some don't, and the law of averages takes care of the rest.
The 10x-ARR conversation usually goes like this:
- Us: "Unpack that for me. How do you plan to get there?"
- Them: "Well, we figured if we did X and Y..."
- Us: "What's the relationship between those activities and revenue?"
- Them: (silence)
- Us: "What's your current CAC? Payback period? LTV? What's your conversion rate at each stage of the funnel?"
- Them: (more silence)
The 10x goal is a wish. It's not connected to inputs. The founder doesn't know what their marketing spend turns into revenue dollars, doesn't know what their unit economics look like, doesn't have a model that connects activity to outcome. They picked the goal because it sounded ambitious in a board meeting.
The frustrating part: the same founders, when asked "what's the relationship between marketing spend and revenue?", often get angry at the question. They want results without economics. They want outputs without inputs. They want the silver bullet but without the seven months of patient forging.
The right line of thinking sounds completely different. "We're in pursuit of better. We want to increase conversion rates. We want more traffic. We're going to make swings, iterate, and see what compounds." That's the gold BB mindset. It's the founder who's actually going to 10x their ARR — by patiently firing a hundred shots until enough of them hit.
We covered the systemic version of this argument in our Marketing for SaaS post, which gets into why "we're focusing on sales" is the same misunderstanding dressed up in different clothes. Both come from founders who want outputs disconnected from inputs.
The tier breakdown:
- Iteration-mindset (great fit). Founder talks about experiments, hypotheses, conversion rates, payback periods. Has a model — even a rough one — that connects activity to outcome. Has realistic expectations for what changes produce what magnitude of result.
- Ambitious-but-grounded (workable). Founder has stretch goals tied to real constraints (e.g., funding milestone, board target). Willing to invest at the level required and accept that the timeline might be longer than ideal.
- Magical-thinking (red flag). Founder wants 10x results with 1x investment, expects step-changes from incremental work, and can't articulate the economic model that would make their goal possible.
The fix: Before you write down a revenue goal, write down the path. What channels, what conversion rates, what unit economics, what timeline, what investment. If you can't reverse-engineer the goal back to inputs, the goal isn't a goal — it's a fantasy. And no agency, marketing team, or amount of money can deliver against a fantasy.
Criterion 5: Taste — three tiers of founder self-awareness
This is Zach's territory, and it's the hardest one to evaluate because taste is partly subjective and partly objective and founders almost always overestimate their own.
The setup: most SaaS founders had to do their own copy, design, branding, and website at some point in their company's life. There was no money or time to hire someone, so they made it themselves. That's fine. Founders make a lot of things they aren't qualified to make. It's how startups work.
The problem isn't that they made it. The problem is what happens next, when they consider whether to change it. There are three tiers:
- Tier 1 (great fit). "I'm not a designer. I'm not a copywriter. I made this because I had to, but I know it's not great. I'd rather allocate my time to the things I'm good at and trust someone else with this." This founder will be easy to work with because they have humility about their own taste and clear deference to expertise they don't have. They get value out of the engagement and we produce work we're proud of.
- Tier 2 (workable). "I tried to make this and it didn't turn out how I wanted. I hired someone before and wasn't thrilled with the result. I know it needs work but I have opinions about what's wrong." This founder is fine to work with — they have judgment, they just haven't found the right execution partner yet. The work is figuring out what they actually want vs. what they think they want.
- Tier 3 (red flag). "I did all the design and copy myself. I think it's great. I'm not interested in changing the look and feel. Just help me with [narrow tactical task]." This founder cannot be helped, because if they thought their existing work was great they wouldn't have reached out to us in the first place. They're seeking validation of work they secretly know is below the bar, while protecting the ego investment they made when they produced it. We can't fix that. We can only decline the engagement.
The deeper thing about taste: there are two skills here that look like one. There's the ability to recognize good work, and there's the ability to produce good work. They develop on different timelines.
Most founders develop the recognition skill faster than the production skill. They can look at two designs and tell you which one is better, even if they can't make either of them. That's actually good — it's the floor for being a good client. They can give meaningful feedback on options we present, even if they couldn't have produced those options themselves.
Founders who lack the recognition skill — who can't reliably tell good design from bad design, good copy from bad copy — are the ones we struggle with most. Not because they're bad people, but because the feedback loop breaks. We can't optimize against a quality bar that the founder can't see.
There's a useful test for this: when we present three design options, the great-fit founder can tell us which one resonates and why. The workable founder can pick one but can't always articulate why. The red-flag founder either picks based on personal preference unrelated to the brief, or insists on combining elements from all three into something that violates basic design principles.
Zach has a discipline he uses to manage his own taste reactions when reviewing junior designers' work. He asks himself: "Am I upset at how they did the work, or am I upset at the result?" If the result is good but the path was different from how he would have done it, he lets it go. That's a hard discipline. It's also what separates founders who can scale a creative team from founders who become bottlenecks on their own company.
For copywriting specifically, the bypass for taste subjectivity is the AB test. A founder who says "I don't love this headline" is making a taste claim. A founder who says "let's AB test this against the alternative and let the conversion rate decide" is making an objective claim. The first is hard to work with. The second is delightful.
The fix: Before you decline an agency's recommendation, ask yourself which version of "no" you're giving. If you're saying no because the work doesn't match your taste, ask whether your taste is calibrated for the outcome you want. If you're saying no because the data suggests a different direction, that's a legitimate signal. Confusing the two is the most common way founders sabotage their own engagements.
Two more criteria, kept private
We have two additional criteria we use internally that we're not going to share publicly. Sorry. They're a competitive moat for us — they let us screen out specific failure modes that have shown up in client work and would be easy for prospects to game around if we publicized them.
What we will say: both of them are subtle behavioral patterns that aren't visible until you've worked with someone for a few weeks. We don't catch them on discovery calls. We catch them once the engagement is underway, and at that point we usually have to either restructure the engagement or end it.
If you want to know whether you'd score well on the private criteria, the best heuristic is: did you score well on the five we just covered? Founders who hit the great-fit tier on impact, async, timelines, expectations, and taste almost always hit the great-fit tier on the private ones too. The criteria aren't independent. They're symptoms of the same underlying thing — strategic clarity, intellectual humility, and a willingness to trust expertise outside your own domain.
The pattern: this is one trait, not five
Looking at the framework as a whole, here's what it's actually measuring. Not five independent traits, but one underlying meta-trait expressed five different ways: the ability to think clearly about your own business, articulate it precisely to other people, and trust expertise you don't have.
That trait produces good behavior on every criterion above:
- Strategic clarity → realistic impact requests
- Articulation skill → async-native communication
- Process understanding → reasonable timeline expectations
- Economic literacy → grounded expectations of results
- Intellectual humility → calibrated taste judgments
Founders who have this trait are rare. They're also disproportionately successful, because the trait is exactly what's required to build a venture-scale company. Founders who don't have it can sometimes succeed anyway, but they leave a trail of frustrated employees, partners, agencies, and vendors behind them — and they pay a tax on every relationship they have with someone whose expertise they need.
If you read all of this and think "okay, I'm not in the great-fit tier on some of these" — that's actually the good news. The criteria are skills, not fixed traits. You can develop strategic clarity. You can practice writing things down. You can build economic literacy. You can calibrate your taste by deliberately exposing yourself to good work.
The founders who get worse over time are the ones who don't recognize the gap. The founders who get better over time are the ones who see the gap and treat closing it as a real priority.
For the mirror version of this post — how to spot a bad agency before you hire them — read our Marketing Agency Pricing post, which covers the red flags on the other side of the relationship. Hourly billing, scope creep, lack of transparency. The patterns are remarkably symmetric.
FAQs
What if I score badly on one of these criteria but well on the others?
Probably fine. The framework predicts engagement quality, not founder quality, and any one criterion has limited predictive power on its own. The pattern that worries us is when founders score poorly on three or more — that's usually when the underlying meta-trait is the issue, and no single fix will resolve it. If you score great-fit on four and middle-tier on one, you'll be a strong client. If you score middle-tier on most of them and red-flag on one, the engagement will work but with friction.
Don't you lose clients by being this picky?
We turn down work, yes. We also have a near-zero rate of bad engagements, and most of our clients refer us to other founders within their first three months. The math works because bad engagements destroy more value than they create — they drain the team, damage the agency's portfolio when the client publishes poor work over our protests, and consume the bandwidth that could have gone to a good-fit client. Saying no to the wrong fit makes saying yes to the right fit possible.
What if I'm a first-time founder and I don't have the strategic clarity to ace this framework?
That's normal and expected. Most first-time founders don't have the meta-skills the framework measures, because the meta-skills come from doing the work. The advice for first-time founders isn't "be like a sophisticated buyer" — it's "find an agency that's good at coaching first-time founders, and budget for the coaching to take time." Some agencies (including us, sometimes) take on first-time founders explicitly because we're willing to be part of the strategic development. But we're explicit about it. We charge for it. And we set the expectation that the first three months will involve a lot of teaching, not just delivery.
How do I know if I'm in tier 3 on taste without realizing it?
The hardest one to self-diagnose, because tier-3 founders by definition don't see their own work clearly. Three diagnostic questions: (1) When you show your website to designers and copywriters you respect, do they offer concrete feedback or politely change the subject? (2) When you A/B test elements of your site, does the version you personally prefer usually win, or does the data surprise you? (3) Have you ever rebuilt your brand or website because you finally admitted it was bad — and if so, how recently? Tier-3 founders never get to that admission. If you've made it before, you're probably already in tier 2 or better.
Should I show this post to my marketing agency?
We'd encourage it. If your agency reads this and says "yes, this matches how we evaluate clients too" — that's a good sign. If they say "we'd never be this picky, we work with anyone" — that's a different signal entirely. The best agency-client relationships are mutual selections. Both sides know what they're looking for and are willing to walk away when the fit isn't there.
What about the two criteria you're keeping private?
We won't tell you. But you can guess: they're not about skills, they're about specific behaviors that emerge under pressure. Almost every founder we've ever ended an engagement with showed these behaviors in the first two months — sometimes in the first two weeks — and we missed the signals because we hadn't formalized them yet. Now we have. That's why they stay private. If you're curious whether you'd trigger them, the answer is probably the same as your score on the public five.
