Skip to content

Managing Customer Growth

Innovation & New Business Proposal - TUHH Institute of Entrepreneurship & Institute of Innovation Marketing, Hamburg · part of my Technology Management MBA · study notes for revision.


Everything in this module so far has been about earning the right to grow. We modelled the business, gathered evidence about customers and rivals, and tested the riskiest assumptions cheaply. This last chapter asks the question that follows: now that people want the thing, how do we get more of them, and how do we know the growth is real rather than bought?

The most useful idea here is that growth is not one activity. It is a sequence with gates in it, and each gate has a different question attached. Pushing on the accelerator before you have passed a gate is the classic way for a funded startup to die quickly instead of slowly. Once you are through, the job becomes narrow and quantitative: know which engine you are running, know the two or three numbers that engine implies, and check that a customer is worth more than they cost to acquire.

1 · When to grow: the three stages of company building

Section titled “1 · When to grow: the three stages of company building”

The module frames company building as three stages, each with its own question, its own fit criterion and its own rough timescale. The stages are not interchangeable, and the whole point is that you do not move to the next one until the current question has an evidence-backed answer.

1 · Business discoveryProblem-solution fit · 3 to 6 months
→
2 · Business validationProduct-market fit · 1 to 2 years
→
3 · Business scalingTraction and growth · 3 to 5 years
Each stage answers one question. Discovery asks whether there is a problem worth solving; validation asks whether we can build something customers want; scaling asks how we accelerate growth.
StageThe question being askedThe fit you are looking forRough time
1 · Business discoveryDo I have a problem worth solving?Problem-solution fit3 to 6 months
2 · Business validationCan I build something customers want?Product-market fit1 to 2 years
3 · Business scalingHow can I accelerate growth?Traction and growth3 to 5 years

Notice how long stage 2 takes. A year or two of validation is the norm, not a sign of failure, and it is precisely the stage that founders under investor pressure try hardest to skip. Spending on acquisition during stage 2 does not create product-market fit; it just buys a larger sample of people who will leave.

2 · Measuring product-market fit: retention is key

Section titled “2 · Measuring product-market fit: retention is key”

The module is blunt about this. The metric that validates the actual value users get from a product is the one that measures repeat usage of the product over time. Everything else is a proxy, and most of the popular proxies are bad ones.

Bad measures of product-market fit
  • Website visitors
  • Conversion rate
  • Registered users
  • Freemium users who are not paying
  • Net Promoter Score
What actually validates it
  • Repeat usage measured over time
  • Retention curves that flatten rather than decay to zero
  • The share of a signup cohort still active months later
  • Evidence that users get real value, not that they arrived

Why are these five bad? Because every one of them can be raised without the product getting any better. Visitors and registered users respond to advertising spend. Conversion rate responds to a redesigned button. Freemium users who never pay have told you nothing about willingness to pay. Net Promoter Score asks people to predict their own future behaviour, which they do badly. A retention curve cannot be faked in the same way: either people came back, or they did not.

The practical form is a cohort retention curve. Group users by the month they signed up, then track what percentage of each cohort is still active one month later, two months later, and so on. A curve that decays towards zero means no product-market fit no matter how impressive the signup numbers look. A curve that flattens out at some level means a stable core of users found real value, and that flat portion is the part of the business you can actually grow.

3 · Conversion rate optimization: the product is a funnel

Section titled “3 · Conversion rate optimization: the product is a funnel”

The second framing is that your product is a funnel or a growth loop, and every step in the funnel has a drop-off. This matters more than it first appears, because the drop-offs multiply rather than add. A funnel with five steps that are each 50 percent efficient delivers about three percent of the people who entered it, not fifty percent.

The compounding works in your favour too. Improving one step from 20 percent to 25 percent does not add five percent to the outcome, it multiplies the whole downstream flow by 1.25. This is why conversion work is attractive: the leverage sits in whichever step is worst, and fixing that one step lifts everything behind it.

A venture does not grow in a general way, it grows through a specific mechanism. The module names three, and the practical instruction is to know which engine you are actually running and to focus on the small number of metrics that engine implies. Running all three badly is the common failure.

Paid engineSpend money to acquire customers. Works only while a customer is worth meaningfully more than they cost. Watch CAC, LTV and the ratio between them.
Sticky engineWin by keeping customers. Growth happens whenever the rate of new customers exceeds the churn rate. Watch churn and retention above all.
Viral engineCustomers bring customers as a side effect of using the product. Watch the viral coefficient and the cycle time.
Three engines, three sets of metrics. The choice is not a preference, it follows from what the product actually is.
EngineHow growth happensThe condition for it to workThe metrics that matter
PaidMoney is spent on acquisition and converts into customersLifetime value comfortably exceeds acquisition cost, with margin left over for overheadsCAC, LTV, LTV/CAC ratio, payback period
StickyCustomers arrive and then stay, so the base accumulatesNew-customer rate is higher than the churn rateChurn rate, retention rate, cohort curves
ViralExisting customers bring new ones through normal useViral coefficient above 1.0, with a short cycle timeInvites sent, invitation acceptance rate, cycle time, viral coefficient

The sticky engine deserves a second look because its arithmetic is unforgiving. If you add customers at 5 percent per month and lose them at 7 percent per month, you are shrinking regardless of how good your marketing is. Every point of churn removed is worth more than a point of acquisition added, because it compounds over the whole base rather than over one month’s intake.

5 · The funnel and matching methods to its stages

Section titled “5 · The funnel and matching methods to its stages”

The module uses a five-stage customer funnel and makes one central point about it: marketing methods must be matched to the stage, not sprayed across the whole thing.

Awarenessthey learn you exist
Interest and engagementthey try, explore, register
Transactionthey buy
Loyalty and repurchasethey come back
Advocacythey bring others
The five stages. Prospects enter at the top not yet ready to transact, and the methods that reach them there are different from the methods that convert them lower down.

The matching rule from the slides is specific:

  • Top-of-funnel prospects, who are not yet ready to transact, are more likely to be reached through display and social media advertising, press coverage and public relations, referrals from social networks and bloggers, and content marketing.
  • Prospects who arrive at the middle of the funnel, ready to transact, are more likely to be reached through search engine marketing, search engine optimization and affiliate marketing.

The logic is about intent. Somebody typing a problem into a search engine has already decided they want a solution, so search methods harvest existing demand. Somebody scrolling a social feed has decided nothing, so the job there is to create demand and awareness. Spending search budget on people with no intent, or content budget on people ready to buy right now, wastes both.

Awarenessdid they notice
Landing page bounce rate, the share who visit and leave immediately
Share of traffic from first-time visitors
App launched as a percentage of downloads
Interest and engagementdid they try it
Share interacting through likes, shares, e-mail signups
E-mail open rate, click-through rate beyond the landing page
Completed onboarding, enrolment or registration rate as a share of visitors
Share of registered users active in a period
Ratio of daily to monthly average users, and average pages per session
Transactiondid they buy
Conversion rate of visitors to purchasers
Free-to-paid conversion rate
Checkout abandonment rate, transaction size
Profit: transaction margin minus campaign cost, and LTV/CAC
Loyalty and advocacydid they stay and tell others
Retention rate, repurchase rate, transactions per period
Revenue per customer per period
Net Promoter Score
Invites sent, invitation acceptance rate, cycle time, viral coefficient

The ratio of daily average users to monthly average users is worth singling out. It is a compact engagement measure: a ratio near 1.0 means your monthly users are essentially daily users, while a ratio of 0.1 means the typical user shows up about three times a month. For a product that claims to be part of someone’s daily workflow, that ratio is a lie detector.

6 · Growth channels: owned, earned and paid

Section titled “6 · Growth channels: owned, earned and paid”

Most companies grow using one or two channels, not all of them. The module sorts digital marketing methods into three categories by who controls the message and who pays for the delivery.

Paid mediayou control it, an agent delivers it for payment
Owned mediayou control both message and channel
Earned mediaa third party relays it, unpaid, because it is worth sharing
The three categories differ in control, in credibility and in cost structure, which is why a venture usually needs more than one.
CategoryWhat it isProsCons
PaidThe message is under the company’s control and delivered by an agent in exchange for payment, as advertisingControl over targeting and message; effectiveness can be measured wellVariable cost per customer acquired
OwnedMessage and delivery channel are both under the company’s control, as with the company blogControl over the message; no variable costsRequires steady effort and coordination of content development, SEO and e-mail marketing; fixed costs
EarnedThe message is relayed by a third party who accepts no payment and thinks it is worth sharingNo variable costs; credibility from third-party endorsementNo control over targeting or message; negative referrals go viral too

Types of owned media online are the company website, the company blog, company social media accounts, company press releases, and e-mail to prospects who have approved contact. Offline equivalents, which the module lists but does not pursue, are trade show booths, brochures and conference presentations.

Types of paid media online are search engine marketing, display and social media advertising, and affiliate marketing. Offline equivalents are magazine, newspaper, radio, television and outdoor advertising, event sponsorship, and direct mail from purchased lists.

Types of earned media are public relations and press coverage, reviews and referrals by influential bloggers, and word-of-mouth referrals from individual consumers through social networks and face-to-face channels.

7 · Working the owned channel: A/B testing, SEO and content

Section titled “7 · Working the owned channel: A/B testing, SEO and content”

A/B testing is how a website gets optimized: show two versions to comparable groups and let the conversion data decide, rather than the loudest opinion in the room.

Search engine optimization starts with understanding what a search engine does. It crawls, so provide a sitemap, including for dynamic content, to make the crawl easy. It indexes, so make sure your site is indexed for the right kinds of keyword. It ranks, and here the number and quality of inbound links into a page signals how authoritative that page is.

Organic search traffic is described as very attractive for two reasons: it is free, and searchers are often primed to transact. The work splits in two:

  • On-page optimization means determining the keywords you are trying to rank a page for, and experimenting with them.
  • Off-page optimization means increasing the number of high-quality inbound links to your site so it ranks higher.

Placement matters enormously. The first page of organic results captures about 89 percent of all clicks, and the first listing on that page takes about 42 percent. Good sources of inbound links are press coverage and content marketing, and free tools exist to track and score who is linking to you.

Content marketing, through blog posts, infographics, how-to videos, short online courses, webinars, widgets, free tools, white papers, e-books and podcasts, demonstrates your domain expertise and delivers value to prospects, and so drives organic search traffic. The caution attached is about timing: content marketing requires upfront and ongoing investment, and although the eventual payoff can be very high, it builds slowly, because prospects acquired this way may not yet be actively searching for a solution and will need nurturing.

8 · Working the earned channel: press and blogs

Section titled “8 · Working the earned channel: press and blogs”

Managing press and public relations comes with a short list of best practices. A compelling story topic is timely, impactful, and full of conflict and human interest. Craft the e-mail subject line carefully. Identify and then cultivate relationships with key reporters in your domain by engaging with their work and sharing insights and story ideas before you pitch your own venture. Personalise the pitch by showing you know their past work and why your story fits what they cover. Reporters at major outlets often track and echo coverage by influential bloggers and specialised outlets, so it frequently makes sense to start small and build press momentum. And follow up after coverage: respond to readers’ comments and forward the story to other outlets who might echo it.

Blogs work similarly. Bloggers with a relevant focus may recommend your product if they are persuaded it will deliver value to their audience. Some request compensation for referrals, which is an accepted practice that can yield a positive return, but deal only with bloggers who comply with the disclosure rules for endorsements in advertising. The honest limitation is that blogger referrals are an effective early tactic that does not scale readily: qualifying and approaching bloggers is time consuming, and most will promote your product only once. So test and optimise the pitch, and concentrate effort on the more influential ones.

A product grows virally when it spreads directly from customer to customer. The module lists four mechanisms, and notes that with the exception of incentives, none of them involves marketing spending.

Direct network effects
Some products only function when used jointly by two or more parties, so users interact directly and each new user pulls in others. Early internet telephony required both the caller and the recipient to have the software.

Word of mouth
A happy customer recommends the product to somebody else. Deliberate, explicit, and driven entirely by whether the product is actually good.

Casual contact
Like a common cold, the product spreads through incidental customer-to-customer contact with no explicit recommendation. The classic case is a free e-mail service that appended a sign-up link to the bottom of every message its users sent.

Referrals through incentives
The company offers a reward that encourages existing customers to recruit new ones. The only one of the four that costs money.

The viral coefficient reflects, for a typical new customer, the number of additional new customers acquired. A coefficient of 1.0 means one new customer attracts exactly one more.

Viral coefficientk = invitations sent per new customer per period × average conversion rate of those invitations
Users after n cyclesTotal = U0 × (1 + k + k² + k³ + … + kⁿ)

Three things follow, and they are the whole practical content of the concept:

  1. A coefficient above 1.0 compounds. Each cycle produces more new customers than the last, which is the explosive growth everybody wants. The module is careful to say that this is rare.
  2. A coefficient below 1.0 produces a decaying series. Each cycle produces fewer additions than the last and the total converges to a finite multiple of the original cohort. That is not useless. A coefficient of 0.5 still means every customer you acquire effectively brings one more over time, which halves your effective acquisition cost, but it will never grow the business on its own.
  3. Cycle time governs the speed. Virality depends critically on the invitation cycle time, meaning the average time a new customer takes to issue invitations plus the average time recipients take to respond and sign up. The same coefficient with a two-day cycle and a two-month cycle produce wildly different curves.

The number of invitation cycles available also varies enormously by product: for some, new customers exhaust most of their invitation opportunities within weeks; for others, customers keep issuing invitations for years. And virality is hard to observe. The most accurate way to measure the coefficient is to survey a sample of new customers and ask what actually motivated them to sign up.

10 · Unit economics: what a customer is worth against what they cost

Section titled “10 · Unit economics: what a customer is worth against what they cost”

The last block turns growth into a financial question. Unit economics are the direct revenues and costs associated per unit of the business model, in the simplest form gross profit per unit minus cost per unit. For customer growth, the two quantities are:

  • Customer acquisition cost (CAC), the cost to acquire a single customer, typically through sales and marketing.
  • Customer lifetime value (CLV or LTV), the total gross profit or contribution a single customer generates over the course of their use of your service, out of which R&D, general and administrative expenses and other overheads must be covered.
Basic lifetime valueCLV = (P × L) - CAC
Retention and churnRR = 1 - CR
Expected purchasing lifetimeL = 1 / (1 - RR) = 1 / CR
The health testCLV / CAC greater than 3
Payback periodPayback = CAC / monthly gross profit per customer

Here P is the periodic profit one customer generates, which may be measured as average gross margin per user, as contribution margin (revenue minus variable cost) or as gross profit (revenue minus cost of goods sold), and revenue can include both one-time and recurring streams. For a one-time revenue stream you additionally need the lifetime of the product and a replacement probability. L is the expected purchasing lifetime in years or months.

The rule of thumb the module states is that CLV divided by CAC should exceed 3 in the long run to cover overheads. The reasoning behind the factor of three is that lifetime value is gross profit, not net: sales and marketing beyond acquisition, product development, and general and administrative costs all still have to come out of it. A ratio of 1 means you break even on the customer and fund nothing else.

Alongside the ratio sits the payback period: how long until a customer has repaid what it cost to acquire them. Unless the business is B2B, the module wants a shorter payback, roughly 6 to 18 months, because a long payback means growth consumes cash faster than it returns it. There are only two ways to shorten it: reduce the CAC, or increase the monthly profit margin per customer.

In a sustainable business, CAC eventually falls below CLV and then levels off, still requiring ongoing investment but costing less than a customer is worth. CLV often rises over time as well, through upselling to existing customers, and pricing power can increase if the product becomes a standard with little forceful competition. The gap before that crossover is the cash burn.

For a segment rather than an individual, the arithmetic is simply scaled: the CLV of a customer segment equals the CLV of the average customer multiplied by the number of customers in the segment. And where the horizon is long, the module incorporates the time value of money, discounting each year’s profit by a net present value factor before summing.

The module closes with three warnings that cut against the instinct to systematise everything immediately.

Do things that do not scale. Startups take off because the founders make them take off. There may be a handful that grew by themselves, but usually it takes some sort of push. The most common un-scalable thing founders have to do at the start is recruit users manually, and founders resist it because they would rather sit at home writing code than go out and talk to strangers and probably be rejected by most of them.

Focus on a single narrow segment. Doing so helps a startup in two ways: it can tailor the product to that segment’s needs and so deliver distinctive value, and it conserves scarce resources, since targeting one narrow segment is less expensive than targeting several or a very broad one. The caveat attached is that while you are still searching for product-market fit, it is fine to test demand with prospects outside your early adopter segment.

Expect a chasm. Visionary early adopters see the promise of a revolutionary new product; mainstream customers require clear performance benefits plus a reputable vendor’s reliable whole-product solution. A chasm separates the two groups because visionaries’ recommendations are not credible to pragmatists, which is why revolutionary products often stall after an initial growth surge. To cross it, a firm can launch an intensive campaign to penetrate one mainstream segment and then target adjacent segments from that beachhead. The means of crossing are listed as positioning (being unique in buyers’ budget considerations), product (augmenting the core into a whole product), pricing, promotion (get, keep and grow customers) and place (delivery channels).

(a) Funnel arithmetic: where the leverage sits

Section titled “(a) Funnel arithmetic: where the leverage sits”

A subscription tool runs a monthly funnel starting from 40,000 visitors. Gross profit per paying customer is 300 euros over their lifetime.

StageConversionPeopleComment
Visitors-40,000Top of funnel
Started signup12%4,800
Completed onboarding45%2,160The worst step
Started free trial70%1,512
Converted to paid20%302

End-to-end conversion is 302 / 40,000 = 0.76 percent, and the funnel is worth 302 × 300 = 90,600 euros per month.

Now improve only the onboarding step, from 45 percent to 60 percent, and leave every other rate untouched:

StageConversionPeople
Visitors-40,000
Started signup12%4,800
Completed onboarding60%2,880
Started free trial70%2,016
Converted to paid20%403

Paying customers rise from 302 to 403, a 33 percent increase, worth an extra 30,300 euros per month. The multiplier is 60/45 = 1.333, and it applies to everything downstream. Buying that same 33 percent uplift through the top of the funnel would have required finding another 13,300 visitors every month, and paying for them forever.

Two scenarios, both starting from a cohort of 1,000 customers.

Scenario Ak = 6 invites × 0.20 conversion = 1.20
Scenario Bk = 4 invites × 0.15 conversion = 0.60
CycleA: new users (k = 1.2)A: cumulativeB: new users (k = 0.6)B: cumulative
01,0001,0001,0001,000
11,2002,2006001,600
21,4403,6403601,960
31,7285,3682162,176
42,0747,4421302,306
52,4889,930782,384

Scenario A accelerates: each cycle contributes more than the last, and the cumulative total has not begun to level off. Scenario B decays and converges towards 1,000 / (1 - 0.6) = 2,500 users no matter how long you wait. Scenario B is not worthless. Each acquired customer eventually brings 1.5 more, so an effective CAC of 100 euros of spend becomes 100 / 2.5 = 40 euros per customer actually gained. But B will never grow the business on its own, so it must be paired with a paid or sticky engine.

Cycle time decides how long these rows take. If a cycle is one week, scenario A reaches cycle 5 in roughly a month. If a cycle is three months, the same five rows take more than a year, by which time competitors have moved.

A venture charges 30 euros per month with a 40 percent gross margin, so periodic profit is 12 euros per customer per month. Monthly churn is 5 percent.

Expected lifetimeL = 1 / 0.05 = 20 months
Lifetime valueCLV = 12 × 20 = 240 euros
Ratio at CAC = 100240 / 100 = 2.4 → below the threshold of 3
Payback period100 / 12 = 8.3 months → inside the 6 to 18 month range

The payback is acceptable but the ratio is not, so the business covers acquisition and a little more, and leaves too little for overheads. Three levers, and it is worth seeing which is strongest. Cutting CAC to 80 euros lifts the ratio to 3.0. Raising the margin from 40 to 50 percent lifts profit to 15 euros and the ratio to 3.0. But cutting monthly churn from 5 percent to 3.5 percent stretches the lifetime to 28.6 months and lifts CLV to 343 euros and the ratio to 3.4, the biggest single improvement of the three, because lifetime enters the calculation as a reciprocal. This is the arithmetic behind the claim that retention is the highest-leverage thing a growing venture can work on.

Basecamp: growing without buying growth. Basecamp is a clean illustration of a sticky engine backed by owned and earned media. The project management product launched in 2004 and grew from over 100,000 accounts in early 2006 to over a million later that year, more than three million by 2009, over fifteen million by 2014, and by 2016 was still adding more than 7,000 accounts weekly, all with headcount under fifty and only one outside investor. The company was profitable from inception, which meant the founders could optimise for long-term product and company health rather than for growth at any cost. What drove acquisition was not paid media. Basecamp was an early adopter of content marketing and of search advertising, but did much less paid acquisition than comparable software ventures, and its visibility came instead from the founders’ blog, best-selling books, contributions to the developer community and regular columns in the business press. Surveys indicated that positive word of mouth about the product’s ease of use carried it between project managers in industries as different as advertising, construction, education, retail and religious organisations. Three design choices reinforced the stickiness: a genuinely no-touch, self-service model that required no sales outreach to place an order, which let a very small team serve a very large base; a deliberate refusal to price per user or per seat, bucking near-universal industry practice, on the belief that it encouraged broader usage across a customer’s team and made the product stickier; and a refusal to build a negotiated enterprise tier, keeping the product tailored to teams of five to twenty people even as large organisations adopted it. The pricing question the case turns on is instructive for this chapter: the team explicitly defined optimal prices as those that maximise lifetime value, and were willing to accept a lower purchase rate if a higher average value offset it, or a lower average invoice if it led to higher retention and therefore greater LTV. The honest difficulty they name is that LTV is hard to test directly, because it is a long-term outcome that is especially sensitive to cancellation rates.

Hubble Contact Lenses: a paid engine under scrutiny. Hubble is the opposite engine and shows exactly where its limits bite. It sold daily disposable contact lenses direct to consumers on a subscription, at one dollar per day, or 50 cents per lens, plus three dollars monthly shipping, bypassing the intermediary markups that made the category expensive. Growth came almost entirely from aggressive digital advertising, with one-click purchase straight from social ads into a mobile-optimised store: 52 percent of customers said they had heard about Hubble through social media, and 90 percent of site visits and 70 percent of sales came through mobile. It worked spectacularly at first, reaching an estimated 20 million dollars of sales and a valuation above 200 million within about eighteen months on more than 30 million dollars of funding. What the case makes visible is the unit economics underneath. Management said only two acquisition metrics truly mattered, CAC and LTV and the ratio between them, and used that ratio to guide spending. Average CAC across channels was about 100 dollars against a monthly media budget averaging 1.5 million dollars. Gross margins were roughly 40 percent, deliberately below the 80 percent-plus common in the industry, and the introductory offer gave away the first 30 lenses at a cost of about 13 dollars including shipping. After that free box, about 90 percent of consumers initiated another order, and those remaining churned at about 4 percent per month, while typical customers ordered only about eight months of supply per year rather than twelve. Put those numbers together and the business is not obviously comfortable: a 40 percent margin on roughly 30 dollars a month for eight months a year, discounted by churn and by the giveaway cost, has to clear a 100 dollar CAC by a factor of three. Two further points from the case matter for this chapter. First, CAC and LTV vary by channel, not just in aggregate: the younger audience on one social platform was more price sensitive and churned faster than customers from other channels, so channel-level LTV/CAC is the right unit of decision, not a blended average. Second, management named the boundary explicitly, saying the challenge is that there is a point past which growth stops making sense, where churn gets so high or CAC grows to a level where it is too expensive to acquire and service a customer. And they admitted the gap: they had not spent much time on improving retention or winning back lost customers. For a company running a paid engine at that scale, that is the most expensive omission available.

  1. Locate yourself honestly on the three-stage map. Write down which question you are actually answering right now: is there a problem worth solving, can we build something customers want, or how do we accelerate. If you cannot show a retention curve, you are in stage 2 and no growth spending is justified yet.

  2. Build one cohort retention curve. Group users by signup month and plot the share still active at one, two, three and six months. Look for a flattening. If the curve heads to zero, stop here and go back to the value proposition, because everything downstream in this chapter will only amplify a leak.

  3. Name your engine. Decide whether the product grows because you pay for customers, because customers stay, or because customers bring customers. Pick one primary engine. If the honest answer is that none of the three describes you, that is a finding worth putting in the report.

  4. Choose the three metrics that engine implies and nothing else. Paid: CAC, LTV, payback period. Sticky: churn, retention, new-customer rate against churn rate. Viral: invites per user, invitation conversion, cycle time. Put those three on a single sheet and update them weekly.

  5. Draw your funnel with real numbers at every stage and find the worst step. Compute the end-to-end conversion, then recompute it with that one step improved, so you can see what the fix is worth before you spend anything on it.

  6. Pick one or two channels, not six. Sort your candidates into owned, earned and paid, and match each to the funnel stage it actually serves. Early on, weight your effort towards the top of the funnel rather than towards optimization.

  7. Estimate CLV against CAC, and stress-test it. Compute expected lifetime from churn, multiply by periodic gross profit, subtract acquisition cost, and check the ratio against 3 and the payback against 6 to 18 months. Then redo it per channel, because a healthy blended ratio can hide one channel that loses money on every customer.

  8. Decide what you will do that does not scale, and name the first narrow segment you will win completely before widening. Write both down, because both feel wrong in the moment and get abandoned without a written commitment.

TermWhat it means in plain words
Problem-solution fitEvidence that the problem you picked is worth solving at all; the goal of business discovery
Product-market fitEvidence that you can build something customers actually want, shown by repeat usage over time
Traction and growthThe scaling stage, where the question becomes how to accelerate rather than whether to
Cohort retention curveSignups grouped by period and tracked over time; the honest test of product-market fit
Conversion rate optimizationImproving the pass-through rate at each funnel step, where gains multiply downstream
Paid engine of growthGrowth bought with acquisition spending; works only while lifetime value comfortably exceeds acquisition cost
Sticky engine of growthGrowth by retention; the base grows whenever the new-customer rate exceeds the churn rate
Viral engine of growthGrowth where customers bring customers as a side effect of using the product
Customer funnelAwareness, interest and engagement, transaction, loyalty and repurchase, advocacy
DAU / MAU ratioDaily average users divided by monthly average users; a compact measure of engagement intensity
Owned mediaMessage and channel both under the company’s control; no variable cost, but real fixed effort
Earned mediaMessage relayed unpaid by a third party who thinks it worth sharing; credible but uncontrolled
Paid mediaMessage controlled by the company but delivered by a paid agent; measurable, with a variable cost per customer
Search engine optimizationGetting found in organic results through on-page keyword work and off-page inbound links
Content marketingPublishing genuinely useful material to demonstrate expertise and earn organic traffic; slow to pay off
Viral coefficient (k)Invitations sent per new customer multiplied by their conversion rate; above 1.0 compounds, below 1.0 decays
Invitation cycle timeHow long one round of invite plus response takes; it governs the speed of viral growth
Customer acquisition cost (CAC)What it costs in sales and marketing to acquire one customer
Customer lifetime value (CLV / LTV)The total gross profit one customer generates over their life as a customer
Churn rate / retention rateThe share of customers leaving, and its complement; expected lifetime is one divided by the churn rate
CAC payback periodHow long until a customer has repaid what they cost to acquire; target 6 to 18 months outside B2B
Unit economicsThe direct revenues and costs per unit of the business model; gross profit per unit minus cost per unit
The chasmThe gap between visionary early adopters and pragmatic mainstream buyers, whom visionaries cannot vouch for
Whole productThe augmented offer a mainstream buyer needs, beyond the core product itself
  1. Name the three stages of company building, the question each one asks, and the fit criterion attached to each.
  2. The slides list website visitors, conversion rate, registered users, non-paying freemium users and Net Promoter Score as bad measures of product-market fit. What single property do all five share that disqualifies them, and what is proposed instead?
  3. Explain why the marketing methods that reach top-of-funnel prospects differ from those that reach middle-of-funnel prospects, and give two examples of each from the module.
  4. Viral coefficient calculation. A product’s new customers each send 5 invitations per period, and 24 percent of those invitations convert. Compute the viral coefficient. Starting from 2,000 users, how many new users arrive in the first three cycles, and what does the cumulative total converge to?
  5. Unit economics calculation. A venture earns 20 euros of gross profit per customer per month and loses 8 percent of its customers each month. CAC is 180 euros. Compute the expected lifetime, the CLV, the LTV/CAC ratio and the payback period, and say whether the business passes the module’s tests.
  6. Why is reducing churn usually a bigger lever on CLV than an equivalent-looking improvement in margin or acquisition cost?

Back to the course overview →.