Growth playbook
Eight mechanics that decide how a company grows, each with a chart, a formula and a worked example. They explain most of what I look at when I assess a business and most of what I work on after investing. All numbers on this page are invented to show the mechanics.
The short version
TL;DR- The idea
- Growth is arithmetic before it is creativity. Small changes in the right place multiply.
- Funnel
- Conversion gains compound. Four stages improved by 20% double the customers.
- Budget
- More spend buys customers at rising cost. Judge scaling by marginal, not average, cost.
- Creatives and sales
- Ads wear out, leads go cold, closers run out of time. All three can be planned.
- Money
- Payback creates a cash dip that deepens with growth. Blended numbers hide real costs.
- Retention
- Churn sets a ceiling on the customer base that no ad budget breaks.
Funnel math
Conversion rates multiply.
Every stage between visitor and customer has a rate. The rates multiply, so modest gains at several stages beat a big gain at one.
A funnel is a chain of probabilities. Each rate is a lever, and the levers multiply.
If 3% of visitors become leads, 35% of leads book a call, 70% of booked calls take place and 25% of calls close, then 0.18% of visitors become customers. Changing any one of those four rates changes the result in proportion.
That has a practical consequence. Improving four stages by 20% each does not add up to 80%. It multiplies to 107%. The company wins about twice as many customers from the same traffic, without spending a euro more on ads.
Gains compound
Customers from the same traffic as one stage after another improves by 20%.
Fig. 11.2 × 1.2 × 1.2 × 1.2 = 2.07. The same mechanism works against a company when several stages get slightly worse.
| Stage | Today | Each stage +20% |
|---|---|---|
| Visitors | Today10,000 | Each stage +20%10,000 |
| Leads | Today300 | Each stage +20%360 |
| Booked calls | Today105 | Each stage +20%151 |
| Held calls | Today74 | Each stage +20%127 |
| Customers | Today18 | Each stage +20%38 |
Worked example
- Ad spend for 10,000 visitors at €1.20 per click
- €12,000
- Customers today
- 18
- Acquisition cost today
- €653
- Customers with each stage +20%
- 38
- Acquisition cost afterwards
- €315
What I do about it
- Measure every rate separately, every week
- Start with the stage furthest below what is normal for that type of business
- Change one thing per stage at a time, then measure again
- Treat small losses at several stages as seriously as one big loss
Diminishing returns
The next euro buys less than the average euro.
More budget in a channel reaches people who are harder to convince. Scaling decisions should look at what the extra customers cost, not at the average.
Every channel has a pool of people who are easy to reach and quick to buy. The first budget finds them.
More budget has to reach people who are less interested, so each extra customer costs more. The average cost rises slowly, because it still contains the cheap early customers. The cost of the extra customers rises faster.
That is why a channel can look healthy on average while new budget is already losing money. The useful number is the marginal acquisition cost: the extra spend divided by the extra customers it brought.
Average and marginal acquisition cost
As monthly spend rises, the marginal cost crosses the affordable limit long before the average does.
Fig. 2Schematic model. Real channels saturate at different speeds; the shape is typical.
Worked example
- €10,000 a month brings
- 100 customers
- Average cost per customer
- €100
- €20,000 a month brings
- 162 customers
- Average cost per customer
- €123
- Cost of the 62 extra customers
- €160 each
What I do about it
- Raise budgets in steps and measure the marginal cost of each step
- Set the affordable cost per customer from the payback target
- Open new audiences, creatives or channels before the current one saturates
- Stop increasing spend where extra customers cost more than they bring
Creative fatigue
Ads wear out.
The more often the same people see a creative, the fewer of them click. A steady supply of new creatives keeps the cost per click down.
How creatives are testedOn social platforms, the same audience sees a creative again and again. The people who were going to react do so early.
After that, the click-through rate falls, each click gets more expensive and the cost per lead climbs. Nothing is broken. The ad is simply used up.
A new creative resets the curve. What matters is not one great ad, but a production rhythm that delivers new hooks and formats before the old ones fade.
Click-through rate over twelve weeks
One creative running unchanged, compared with a new creative every three weeks.
Fig. 3Schematic. How fast a creative wears out depends on audience size, budget and format.
Worked example
- Cost per 1,000 views
- €10
- Average CTR, no new creatives
- 0.74%
- Cost per click
- €1.35
- Average CTR, new creative every 3 weeks
- 1.07%
- Cost per click
- €0.93
The same budget buys about 45% more clicks.
What I do about it
- A weekly plan for new creatives, set before results drop
- Frequency and click-through rate tracked per creative
- Winning hooks carried into new formats
- A backlog of tested ideas, so production never starts from zero
Speed to lead
Interest fades by the minute.
The same lead is far easier to reach right after signing up than a day later. Fast follow-up is one of the cheapest ways to get more sales conversations from the same budget.
A sign-up is a moment of attention. The person has the problem on their mind and the page still open.
An hour later they are in a meeting. A day later they have compared three competitors or forgotten the form. Every hour of delay turns paid leads into cold contacts.
Fast follow-up costs little: an automatic confirmation with a booking link, and a person who calls while the lead is still warm.
Chance of reaching a lead, by response time
Index: reaching the lead within five minutes = 100.
- Under 5 minutes100
- 5 to 30 minutes70
- 30 min to 2 hours45
- 2 to 24 hours25
- Over 24 hours12
Fig. 4Schematic index. The exact curve differs by business and offer; the direction does not.
Worked example
- Leads per month
- 300
- Reached when called within 5 minutes
- 60%
- Reached when called the next day
- 35%
- Conversations: fast vs slow
- 180 vs 105
- Extra conversations per month
- 75
Assumed rates. The ad budget is the same in both cases.
What I do about it
- Automatic confirmation with a booking link within seconds
- A setter who calls within minutes during business hours
- Routing rules, so no lead waits for the right person
- Speed to lead reported every week, like revenue
Sales capacity
When marketing works, sales becomes the bottleneck.
Raising an ad budget takes a day. Hiring and training a closer takes weeks. Capacity has to be planned before the leads arrive.
How setters and closers work togetherFor offers sold in a call, every closer has a limit on how many conversations fit into a month.
As long as demand stays below that limit, more leads mean more customers. Once demand passes it, extra leads wait, get rushed calls or are never called back. The cost per lead stays the same, but the cost per customer rises, because part of the paid demand is wasted.
The fix is arithmetic: leads, booking rate and show rate give the calls per month. Calls per closer give the team size.
Demand for calls against closer capacity
With a 35% booking rate and a 70% show rate. Each closer can hold about 80 calls a month.
Fig. 5Example numbers. Calls per closer depend on call length, preparation and follow-up.
Worked example
- 600 leads × 35% × 70%
- 147 calls
- Capacity per closer: 4 calls × 20 days
- 80 calls
- Closers needed
- 2
- 1,200 leads: calls per month
- 294 calls
- Calls lost with only 2 closers
- 134 a month
What I do about it
- A capacity plan per closer, updated with the lead forecast
- Hiring that starts before demand reaches capacity
- Setters who qualify leads, so closers only take calls that fit
- Budget increases tied to available sales capacity
Blended vs paid CAC
The average hides the real cost of growth.
Blended acquisition cost includes customers who came for free. It looks better than the cost of the next paid customer, and the paid cost is the one that matters for scaling.
Many companies divide their ad spend by all new customers. That number is easy to calculate and almost always too flattering.
Part of the customers came through referrals, search or the founder’s network. They would have come without the ads. When the company scales, it is the paid share that grows, so the blended number moves towards the paid number, and the margin shrinks.
Budgets should be planned with the paid acquisition cost per channel. The blended number is useful for the overall P&L, not for scaling decisions.
One month, 100 new customers, €9,750 ad spend
Where the customers came from, and what each calculation says.
- Blended today€98
- Blended after scaling€128
- Paid€150
Fig. 6Bars on a scale from €0 to €200. “After scaling”: 200 paid customers at the same paid cost, organic unchanged at 35.
Worked example
- €9,750 ÷ 100 customers
- €97.50 blended
- €9,750 ÷ 65 paid customers
- €150 paid
- Scale to 200 paid customers
- €30,000 spend
- €30,000 ÷ 235 customers
- €127.66 blended
What I do about it
- Paid acquisition cost reported per channel, separately
- Sources tagged in the CRM, from first click to closed deal
- Budgets planned with paid cost, not blended cost
- Organic and referral growth tracked as its own channel
Cash curve
Growth costs cash before it makes cash.
When customers pay back over several months, every new customer first ties up money. The faster a company grows, the deeper that dip.
Use of funds and runwayEvery new customer costs money on day one and earns it back month by month.
One group of customers acquired in the same month draws a small J-shaped curve: down at the start, up once it has paid back. A company acquires a new group every month, so the curves stack. Growing faster means larger groups entering at the bottom of their curve, and the dip gets deeper and lasts longer.
This is the honest reason a business with good unit economics can still need capital. Capital funds the dip. Shorter payback makes the dip shallower. Slower growth does too, at the cost of speed.
Cumulative cash from acquiring customers
€150 acquisition cost, €40 contribution margin per customer and month, 5% monthly churn.
- Steady: 100 new customers a month
- Growing: new customers +15% a month
- Cash tied up while growing
Fig. 7Only acquisition and contribution margin. Fixed costs come on top.
Worked example
- Steady: lowest point
- −€37k after 5 months
- Steady: back above zero
- after 10 months
- Growing 15% a month: lowest point
- −€66k after 8 months
- Growing: back above zero
- after 15 months
What I do about it
- Model the cash need from payback and growth rate before setting a budget
- Size capital to the dip plus a buffer, not to a round number
- Shorten payback first: upfront payment, annual plans, early upsells
- Watch the low point, not only the monthly result
Customer ceiling
Churn sets a ceiling.
With a steady number of new customers per month, the customer base stops growing when as many leave as join. The level is simple to calculate.
How I read retentionEach month a share of customers leaves. As the base grows, that share is a larger number of people.
At some point the customers who leave equal the customers who join, and growth stops, even though marketing still works. Spending more on acquisition raises the ceiling only in proportion. Lowering churn raises it for free.
Active customers with 100 new customers a month
Three churn rates, four years.
Fig. 8Dashed lines: the ceiling for each churn rate. The base approaches it but never passes it.
Worked example
- 100 new a month at 10% churn
- 1,000
- 100 new a month at 5% churn
- 2,000
- 100 new a month at 3% churn
- 3,333
- Churn from 5% to 3%
- +67% ceiling
Without a single extra euro for acquisition.
What I do about it
- Churn measured by cohort and by reason, not only in total
- Work on onboarding and the first weeks, where most churn starts
- Sales promises checked against what delivery can keep
- Upsells and renewals planned as part of the offer
Summary
All eight on one page.
The number to watch and the main lever for each mechanic.
| Mechanic | Number to watch | Main lever |
|---|---|---|
| 1 Funnel math | Number to watchConversion rate per stage | Main leverFix the weakest stage first |
| 2 Diminishing returns | Number to watchMarginal acquisition cost | Main leverScale in steps, open new audiences |
| 3 Creative fatigue | Number to watchCTR and frequency per creative | Main leverA rhythm for new creatives |
| 4 Speed to lead | Number to watchTime to first contact | Main leverAutomation plus setters |
| 5 Sales capacity | Number to watchHeld calls against closer capacity | Main leverHire before the leads arrive |
| 6 Blended vs paid CAC | Number to watchPaid cost per channel | Main leverClean source tracking |
| 7 Cash curve | Number to watchPayback period and cash low point | Main leverCapital sized to the dip |
| 8 Customer ceiling | Number to watchMonthly churn | Main leverOnboarding and delivery |
These mechanics are not specific to one industry. They apply to software, brands, services and hybrid businesses alike, with different numbers.
Continue reading
More about how I invest.
Next step
Tell me what you are building.
A few sentences are enough: what the business does, where the digital part sits and what you need. A reply on whether a call makes sense follows.