It suits you if
You have a problem that is clear at first sight, a prototype or a service that can be sold, time to run interviews and follow up with customers, and a limited budget for learning.
A marketing plan for startups
A marketing plan for a startup before scaling spend
A practical 12-week framework that helps the team prove the problem and the offer and test the first acquisition channels before marketing spend becomes a large commitment.
Who is the plan for?
Suited to a stage of structured learning, not confirmed scale
This plan works for a startup with a team able to run small experiments that wants to reduce guesswork before raising the budget.
It does not suit you if
The customer or the problem has not been defined at all, the product cannot deliver its core value, or you need immediate growth to cover a severe cash shortfall.
Before you start
Decide who owns the offer decision, who runs the interviews, and who reviews the data weekly. Marketing does not make up for a lack of accountability or slow product development.
Assumptions and limits
What does the plan assume, and what does it not promise?
The plan assumes a reachable market, a problem worth investigating, and the ability to deliver an initial solution. It does not assume that fit has been achieved, and it does not guarantee sales, funding, or a specific acquisition cost.
The core goal
Gather enough evidence to make a decision: continue with the same segment and offer, adjust, or stop before draining larger resources.
Test scope
One core segment, one problem, one initial offer, and two channels at most in each cycle so the result can be interpreted.
Measurement limits
Small numbers do not prove the whole market, but they reveal a direction. Separate free interest, purchase intent, actual payment, and repeat use.
Proving the problem and the offer
Start with evidence before ad messages
Write a falsifiable hypothesis: who is the customer? In what situation does the problem appear? What is its impact? What is the current alternative? And why might they pay for a different solution?
Problem signal
The problem recurs in the customers’ own words, there is a clear cost, delay, or risk, and there were earlier attempts to solve it, even if manual.
Offer signal
The customer understands the outcome, the mechanism, and the limits of the service, and accepts a step with real cost: payment, a booking, a serious trial, or a commitment of time and data.
Warning sign
General liking or a free sign-up is not enough. If customers do not move to a serious step, review the problem, the segment, or the offer before increasing reach.
Customer interviews
Learn from past behavior, not from politeness
Aim for 15 to 25 short interviews spread among potential customers, those who tried alternatives, and those who rejected the offer. Ask about the last time the problem occurred, what they did, who took part in the decision, and what cost them time or money.
Useful questions
When did the problem last appear? How did you try to solve it? What was the hardest step? What made you start looking? And what is stopping the decision now?
Weak questions
Do you like the idea? Will you use the product? How much might you pay? Hypothetical questions encourage polite answers that do not reflect a real decision.
Output of the interviews
A map of customer phrases, purchase triggers, objections, alternatives, and qualification criteria. Use it to shape the offer, the page, and the content.
First channels
Test limited channels that can be interpreted
Choose one existing-demand channel and one direct-reach channel. For example: search or specialized communities to discover demand, with founder outreach or partnerships to reach deeper conversations.
An intent channel
Organic or paid search, a specialized directory, an app store, or a platform the customer turns to when looking for a solution.
A direct learning channel
Founder outreach, referrals, partnerships, small events, or a specialized community. The goal is high-quality conversations, not reach volume alone.
The testing rule
Hold the segment and the offer constant as far as possible, and change one major element at a time. Review the ad testing approach before producing dozens of versions with no hypothesis.
Trust content and the conversion page
Help the customer understand the decision before asking for it
At the start you do not need a huge content library. You need material that answers decision questions, shows you understand the problem, and explains what happens after contact.
Trust content
A guide that explains the problem, a comparison of alternatives, a frank answer on limits and risks, and a documented experience story or a clearly sourced illustrative example.
The conversion page
A headline that names the customer and the outcome, an explanation of the problem and the offer, proof suitable for the stage, FAQs, and one clear step such as requesting a trial or booking a conversation.
Follow-up
Record the source of every opportunity, the reply time, the reason for acceptance or rejection, and the final outcome. Do not settle for counting forms or clicks.
Testing budget
Buy learning before you buy volume
Divide the budget into experiments with a specific question, a maximum loss, a duration, and a decision set in advance. Do not start with a fixed percentage of revenue before you know the unit economics and the team’s ability to serve demand.
Reach and offer tests
Small audiences, messages, and pages to discover which combination generates conversations or qualified requests.
Content and conversion assets
One optimized page, proof materials, and the necessary follow-up and measurement tools. The percentages are indicative and are adjusted to the model.
Learning reserve
Re-testing a promising hypothesis or fixing a fault in the page or measurement. The reserve is not spent just to complete the month.
Decision conditions
When do we scale and when do we stop?
Write the conditions before seeing the results so the decision does not become a justification for earlier spend.
We scale when
Sales or qualified opportunities repeat from the same segment, quality stays stable, operations can serve the demand, and acquisition cost approaches the acceptable limit.
We adjust when
There is clear intent but an objection, the page, or follow-up blocks conversion, or a sub-segment succeeds more than the original hypothesis.
We stop when
The agreed test sample ends with no buying signals, the problem repeats in the unit economics, or success depends on a discount that cannot be sustained.
Measurement indicators
A small dashboard linking learning to economics
Review the indicators weekly in the context of sample size and data quality. Do not use a single indicator to judge the whole path.
Market proof
The number of qualified conversations, the share who describe the same problem, the share who move to a trial or offer, and the number of paying customers.
Path efficiency
Conversion rate, cost per qualified opportunity, response time, attendance rate, close rate, and reasons for loss.
Unit economics
Revenue and margin per customer, acquisition cost, payback time, usage or retention, and the operational capacity to deliver.
12-week plan
From hypothesis to the next quarter’s decision
The plan runs in four phases, each three weeks long, with a short weekly review and a clear decision at the end of each phase.
Understand the problem
Define the segment, run 8–12 interviews, review alternatives, draft the hypothesis, and set success indicators and budget limits.
Build the offer
Complete the interviews, draft the first offer, create the conversion page and trust materials, and set up recording of sources and results.
Run the experiments
Two channels at most, limited messages, daily follow-up of opportunities, and a weekly review of quality, not clicks alone.
Lock the decision
Re-test the strongest hypothesis, calculate the unit economics, document what we learned, and build the next quarter’s cycle according to quarterly planning.
Using an AI assistant
Use it to summarize interviews, collect objections, and produce drafts, with human review of decisions and claims. See the practical getting-started guide.
The final output
A written decision: limited scaling, adjusting the segment and offer, or stopping the hypothesis. Attach the evidence, the constraints, and the next question.
Responsibilities and risks
Who owns each part? And what could spoil the test?
Leading the learning
Attends some of the interviews, settles the offer and the price, and makes sure product and operations decisions respond to what we learn.
Test design
Builds the messages, page, channels, and measurement, and documents hypotheses and results without inflating their significance.
Quality of opportunities
Follows up quickly, records objections and reasons for loss, and feeds the data back to marketing and product.
Changing too many things
Changing the segment, offer, price, and channel together makes it impossible to know the cause of the result.
Illusory success
A discount or a personal network can create first sales that do not repeat. Test repeatability outside the closest circle.
Scaling before operations
Raising spend before follow-up and delivery are ready can raise complaints and hide the real demand signal.
FAQ
Before executing the plan
How large should the testing budget be?
Enough to produce a readable sample within your channel and market, with a loss limit that does not threaten the company’s continuity. There is no general figure that suits every model.
Should I start with ads?
Start with them if there is demand that can be reached, an offer that can be tested, and a page and follow-up. Otherwise, interviews and direct outreach may be faster for learning.
What if we do not get sales?
Review the sequence: did we reach the segment? Was the offer understood? Did serious intent appear? Did follow-up break down? Then decide on one adjustment or stopping.
Are 12 weeks enough?
Enough for a cycle of learning and decision, not for proving the whole market. Models with a long sales cycle need interim indicators and a longer period.
Can the plan be copied as is?
No. The framework is general, while the segment, budget, channels, and decision criteria are built from the reality of the product, the market, and operating capacity.
When do we need outside help?
When the team finds it hard to separate hypotheses, set up measurement, build the page and campaigns, or read results without bias.
Customization note
Plan reviewThis is a starting point, not a final plan for your company
The plan varies with the product type, the sales cycle, profit margin, market, stage, funding, and operating capacity. Do not carry over the percentages, channels, or scaling conditions before replacing them with your actual numbers and constraints.
Turn the hypotheses into a testing program that fits your company
Share your product, market, current stage, and the best evidence you have of the problem or the demand, so we can define what needs proving before scaling spend.
Start a conversation on WhatsApp