Founder Field Note: Our 90-Day Plan Prompt Was Free Management Consulting

The prompt sounded harmless. That was the problem.

A few years ago, we were hiring a Head of Customer Success.

The company was small enough that everyone still pretended the messy CRM was “temporary,” which is founder-speak for “load-bearing spreadsheet.” Churn was higher than we wanted. Onboarding was inconsistent. Sales promised custom miracles. Support knew where the bodies were buried but had stopped writing incident notes because the notes were depressing.

So we did what many hiring teams do when they are underprepared and overconfident: we asked finalists to prepare a “simple 90-day plan.”

Not a huge thing, we told ourselves.

Just a few slides.

Just enough to see how they think.

Just enough to compare candidates.

This is how the dignity tax gets dressed up as rigor. Nobody on the team said, “Let’s extract unpaid management consulting from people who do not work here.” We said, “We need signal.” Then we handed candidates a vague business problem, withheld the actual context, and judged how elegantly they solved our fog.

One candidate, Maya, asked the right questions:

• What segments are churning fastest?
• What does onboarding success currently mean?
• Who owns renewals versus expansion?
• Are there customer health scores, or is everyone reading tea leaves in Slack?
• What decisions would this role actually control in the first 90 days?

The team got uncomfortable. Not because the questions were bad. Because they were accurate.

We wanted a polished 90-day plan. She wanted the operating truth.

Guess which one is more useful.

Takeaways

• A first 90 days interview question is not automatically abusive, but it becomes a trap when the company asks for specificity while giving you guesses.
• If the prompt requires internal data, customer segmentation, pricing context, org politics, or product roadmap knowledge, it is no longer “just a work sample.”
• Your job is not to become their unpaid interim executive so they can decide whether you are worth a calendar invite with Legal.

Maya gave us a diagnostic plan. We secretly wanted a fantasy plan.

Maya did not show up with a 28-slide temple to herself.

She brought a five-page diagnostic plan. It had three phases:

1. Days 1–15: Instrument the truth. Audit onboarding drop-offs, churn reasons, escalation patterns, and customer promises made during sales.
2. Days 16–45: Fix the repeatable leaks. Identify two onboarding bottlenecks, standardize renewal risk language, and create a weekly customer health review.
3. Days 46–90: Scale what works. Build manager dashboards, define ownership between CS and Sales, and publish a playbook for expansion signals.

It was good. Annoyingly good. The kind of good that makes a messy team start adjusting its posture.

But after she left, the debrief turned into one of those quiet corporate séances where everyone summons a concern from the carpet.

“She was thoughtful, but maybe too cautious.”

“I wanted more ownership.”

“Could she move fast enough?”

“I didn’t feel the same energy as the other candidate.”

There it was: the culture fit interview fog machine warming up in the corner.

What we meant was: she did not validate our preferred fiction. We wanted someone to say, “In 90 days, I will reduce churn by 20%, implement health scoring, align Sales and CS, and teach the CRM to love itself.”

Maya said, in professional language, “I need to know what is broken before I prescribe surgery.”

That should have been the green flag.

Instead, our hidden interview scorecard rewarded certainty over judgment. We confused a fantasy roadmap with leadership. Very common founder disease. Symptoms include vague prompts, heroic expectations, and saying “high agency” when you mean “please inherit our operational landfill.”

Takeaways

• If they punish you for asking diagnostic questions, the role may require clairvoyance, not competence.
• “Too cautious” often means “did not pretend our missing information was enough.”
• A strong 90-day answer should show action, but it should also name assumptions. Mature operators use assumptions. Magicians use smoke.

The ugly part: we used her thinking after rejecting her.

Here is the part that still makes me want to crawl under the conference table and live with the abandoned dongles.

We did not hire Maya.

We hired someone who gave a more confident plan. Cleaner slides. Stronger declarations. More executive sparkle. The answer had the Signal-to-Theater Ratio of a TED Talk wearing Allbirds.

Three months later, churn was still ugly. The new hire was not bad, by the way. This is not a revenge fairy tale. They were smart. They were also dealing with the same garbage context Maya had correctly identified.

Then, during a leadership offsite, someone pulled up “that candidate’s framework.”

Maya’s framework.

We did not copy her slides wholesale. That would be cartoon villain behavior. But we used the bones: audit churn reasons, define onboarding success, clarify renewal ownership, build a weekly health review.

In other words, her “insufficiently bold” plan became our actual operating plan once reality showed up with a clipboard.

This is the quiet humiliation candidates rarely get to see. You get rejected for not performing confidence, then your unpaid take-home assignment quietly becomes internal strategy. No credit. No offer. No consulting fee. Maybe not even a rejection email with your name spelled correctly.

The candidate screening process calls this “evaluation.” Candidates experience it as donating expertise to a company that may already know it is underbuilt, underbriefed, or unsure whether it wants to hire at all.

Takeaways

• If your interview work could be used by the team on Monday, you are producing value, not merely proving skill.
• Free consulting interview tasks often hide behind phrases like “show us how you think,” “come prepared with recommendations,” and “walk us through your approach.”
• You are allowed to protect the proprietary layer: exact recommendations, detailed roadmap, implementation sequence, customer-specific analysis, and internal templates.

The better answer is not “refuse everything.” It is “control the altitude.”

Some candidates hear stories like this and decide never to do a take-home again. Fair. Emotionally correct. Sometimes strategically expensive.

The better move is to control the altitude of your answer.

Think of your work sample in three layers:

Layer 1: Proof you have done this before

This is safe and powerful. Use proof blocks from your own history:

• “At my last company, onboarding completion was at 62%. I mapped drop-offs by segment, rebuilt the kickoff checklist, and moved completion to 81% over two quarters.”
• “I inherited renewals with no shared risk language. I introduced red/yellow/green criteria and cut surprise churn in enterprise accounts by 30%.”
• “I aligned Sales and CS by creating a handoff checklist tied to contract promises, not vibes.”

This is your evidence. Bring receipts.

Layer 2: General approach for their situation

This is where you show thinking without giving away the blueprint:

• “Given the limited information, I would start with a diagnostic across churn reasons, onboarding milestones, and ownership gaps.”
• “I would avoid promising a churn reduction target before validating baseline data.”
• “My first priority would be identifying which customer segment is leaking and whether the leak is product fit, implementation, expectation-setting, or support load.”

This is useful but not exploitable enough to become their full plan.

Layer 3: Specific operating plan for their company

This is the dangerous layer:

• Exact segmentation strategy
• Dashboard design
• Customer health scoring model
• Renewal playbook
• Org redesign
• Pricing or packaging recommendations
• Detailed first 90 days execution calendar

That layer should require either employment, a paid work trial, or a much more bounded live working session.

If you are preparing for an AI interview screen or a one-way video interview where the bot asks for a first-90-days plan with no context, use a tool like NoSweatKing to decode the question and turn your real experience into bot-readable answers in your own voice without handing over a full consulting deck.

Takeaways

• Give them proof. Give them approach. Do not casually give them the implementation manual.
• A role-evidence map helps you decide which past examples to share before you start inventing free strategy for strangers.
• If they want Layer 3, ask for scope, time cap, evaluation criteria, and whether the work is paid.

The script Maya should have used — and the script you can use now

Maya did nothing wrong. Let’s be clear. She behaved like a senior operator in a room that wanted a motivational poster.

But candidates should not have to rely on hiring teams being ethical, organized, or awake. You need language that keeps you collaborative without opening your laptop veins.

Here are scripts you can steal.

When they ask for a 90-day plan before sharing context

“Happy to prepare a 90-day view. To keep it useful and fair, I’ll frame it as assumptions, diagnostic priorities, and decision points rather than a detailed operating plan. If you want company-specific recommendations, I’d need churn data, segment context, current ownership, and success metrics.”

This says: I am not difficult. I am accurate.

When the assignment looks like free consulting

“I can share how I would structure the problem and examples of similar work I’ve done. I want to avoid producing a detailed strategy that could be implemented outside the interview process. Would a high-level approach plus a live discussion meet the evaluation need?”

This says: I see the game board.

When they say, “It should only take two hours”

“Great — can you confirm the expected time cap, the scoring criteria, and what level of detail you’re looking for? I’ll keep my submission within that boundary.”

This forces the hidden interview scorecard out of hiding. Or at least makes it sweat.

When they want another round after the take-home

“I’m excited to continue, and I’d like to understand the remaining process. How many rounds are left, who is involved, and what decision still needs to be made?”

Endless interview rounds thrive in polite uncertainty. Make the process name itself.

Takeaways

• Boundary scripts work best when they offer an alternative, not just a refusal.
• Ask for scoring criteria before doing unpaid work. If they cannot define success, they are not evaluating. They are browsing.
• The phrase “to keep it useful and fair” is your friend. It frames boundaries as quality control, not attitude.

How to tell whether the 90-day prompt is legitimate

Not every hiring exercise is a scam wearing business casual. Some teams genuinely need to see how you think. A good process can include a case discussion, portfolio presentation interview, or practical working session.

The difference is whether the exercise respects the candidate as a person with finite time and marketable expertise.

Use this quick read.

Green flags

• They share the evaluation criteria in advance.
• The prompt uses a fictionalized or sanitized scenario.
• The time expectation is clear and reasonable.
• They accept a high-level approach instead of demanding a full roadmap.
• They offer a paid work trial for deep company-specific work.
• The exercise is discussed live, with space for questions and tradeoffs.
• They explain how many steps remain after the exercise.

Red flags

• The prompt asks for a detailed plan using their actual product, customers, funnel, or market.
• They refuse to answer clarifying questions.
• They say “no more than two hours” but ask for work that takes eight.
• They want implementation-ready artifacts: dashboards, playbooks, sequences, pricing analysis, or campaign calendars.
• The process keeps expanding: recruiter screen, hiring manager, panel, take-home, presentation, “casual chat,” executive round, final calibration, vibes with the office plant.
• They call your boundary a lack of passion.

That last one is the cigarette burn on the table. If protecting your time makes you “not passionate,” imagine what working there does to weekends.

Takeaways

• Legitimate assessment tests judgment. Exploitative assessment extracts deliverables.
• A live working session is usually safer than an unpaid take-home assignment because you can show thinking without leaving behind a polished asset.
• Passion is not measured by your willingness to donate strategy to a company with a branded hoodie budget.

Build the “bounded 90-day answer” before you need it

If you are interviewing today, do not wait until the request arrives. Build a reusable answer now.

Use this structure:

1. Your headline

“My first 90 days would focus on diagnosing the biggest customer leak, stabilizing the repeatable process, and creating a measurable operating rhythm.”

2. Your assumptions

“I’m assuming churn, onboarding, and handoff quality are the main risks based on the role description. I’d validate that before committing to targets.”

3. Your proof block

“I’ve done similar work before: at [company], I found that enterprise onboarding delays were driven by unclear technical ownership. I rebuilt the handoff process, added risk flags, and reduced implementation delays by [metric].”

4. Your phased approach

“Days 1–15: listen and audit. Days 16–45: fix the highest-friction repeatable issue. Days 46–90: scale the operating rhythm and define ownership.”

5. Your boundary line

“With more internal data, I could make this more specific. For the interview, I’m keeping it at the approach level so we can discuss tradeoffs rather than pretend certainty.”

That answer is senior. It is structured. It is not submissive. It gives them enough to evaluate you without giving them your whole operating brain in a PDF.

Takeaways

• Prepare a bounded version of the first 90 days interview answer before a recruiter asks for it.
• Tie the answer to past results, not imaginary promises.
• Name what you would validate. Senior candidates do not confuse confidence with guessing loudly.

The founder lesson: candidates are not a free strategy bench

The hiring team has power. That means the hiring team has responsibility.

If a company needs real strategic work, it should pay for it. If it needs to evaluate judgment, it should create a bounded exercise with clear criteria. If it wants a detailed plan for an actual business problem, it should understand that candidates are not a temporary consulting firm with worse snacks.

Maya was good enough. More than good enough. Our process was just built to reward the wrong signal.

That is the part candidates need to remember when a hiring ritual makes them feel small. Sometimes you are not failing the assessment. Sometimes the assessment is measuring obedience, clairvoyance, and slide polish while the real job requires judgment, pattern recognition, and the courage to ask why the CRM is full of haunted data.

Protect your proof. Protect your time. Protect the layer of your expertise that becomes valuable the moment someone can use it without you.

You are not being difficult.

You are refusing to let “show us how you think” become “please fix our company for free.”