# The average family doesn't exist > Klasna's dashboard says the typical learner is 15 and takes about six lessons a month. The customer export shows almost nobody like that, and two very different groups hiding behind the average. - Course: [Marketing Strategy: Who to Win and How](https://productovi.com/courses/marketing-strategy.md) - Module: Who are our customers, really? - Lesson 5/32, ~27 min - Source: https://productovi.com/courses/marketing-strategy/the-average-family-doesnt-exist ## A typical learner **Alex Tutor:** Module 1 gave the plan its first row: Klasna competes mostly with free videos, waiting and school teachers, not with a big school. You're still on the marketing team, and row two asks who Klasna's customers actually are. **Uliana:** Leo's dashboard already answers that: the typical learner is 15 and takes about six lessons a month. So we aim at parents of 15-year-olds and sell a six-lessons-a-month plan. **Sofia:** That sounds like a clear target, at least. **Alex Tutor:** Let's check it against the rows. Here's the sample export: 160 accounts whose first lesson fell between September 2025 and June 2026. **Dataset (CSV):** [Klasna learner accounts, 2025–2026 (sample of 160)](https://productovi.com/datasets/marketing-strategy/the-average-family-doesnt-exist/klasna-customers-2025-2026.csv) A random sample of 160 learner accounts whose first lesson fell between September 2025 and June 2026, exported on 30 September 2026. One row per account. Money in UAH (EU families pay in EUR; the file shows the UAH list price of what they bought). - `customer_id`: Account id, in order of the first lesson - `payer`: Who pays and decides: parent or self (the learner) - `learner_age`: Learner age from account data; used to describe groups, never to aim ads at minors - `country`: ua (Ukraine) or eu (an EU country) - `subject`: english, maths or programming - `goal`: Goal from the sign-up form: national_exam, school_grades, coding_hobby, career, relocation, speaking, keep_up_abroad - `first_lesson`: Month of the first lesson (YYYY-MM) - `last_lesson`: Month of the last lesson; 2026-09 means still active at the export - `active_months`: Months with at least one lesson (summer pauses not counted) - `lessons_total`: Lessons taken - `packages_bought`: Packages bought - `package_size`: Lessons per package: 8, 16 or 32 - `paid_uah`: Paid in total, UAH (packages × size × list price per lesson) - `club_member`: yes if also a Klasna Club member - `heard_from`: First heard of Klasna from: school_teacher, friend_or_family, search, social_network, creator - `considered_instead`: Answer to "what else did you consider?": school_teacher, free_videos_ai, wait_or_nothing, private_tutor, local_centre, large_online_school, none_named **Alex Tutor:** A learner's pace is `lessons_total / active_months`. Now count the accounts that match the dashboard: aged 15 and taking 5 to 7 lessons a month. > **Exercise:** How many of the 160 accounts have `learner_age` = 15 **and** a pace of 5 to 7 lessons per active month (`lessons_total / active_months`, inclusive)? ## Two shapes inside one average **Sofia:** Zero. So the typical learner doesn't exist? **Alex Tutor:** An average blends groups that behave differently. Filter `goal = national_exam` and look at `first_lesson`: some exam families started in the autumn, others only in January to March. **Uliana:** The ones who started after New Year look much busier. Two, three lessons a week? **Alex Tutor:** Check it. Then compare with the younger learners, who mostly take about one lesson a week. > **Exercise:** For accounts with `goal = national_exam` and `first_lesson` of 2026-01 or later, what is the **mean pace** (the average of each account's `lessons_total / active_months`)? Two decimals. **Alex Tutor:** Now the other half of the exam families: the ones who started before New Year. > **Exercise:** For accounts with `goal = national_exam` and `first_lesson` of 2025-12 or earlier, what is the **mean pace** (the average of each account's `lessons_total / active_months`)? Two decimals. ## Spring crammers **Sofia:** Nearly ten lessons a month. But for how long? **Alex Tutor:** Look at `last_lesson`. Of those 40 accounts, 15 stopped in May and 13 in June, right around the exam. Twelve show 2026-09: they're back for the new school year. **Uliana:** Hanna always says 30% of spring families come back in September. Does our sample agree? **Alex Tutor:** Work it out. If the sample disagreed with finance's number, we'd want to know why before writing the plan. > **Exercise:** Of the 40 spring exam accounts (exam goal, first lesson 2026-01 or later), what percentage were active again in September 2026 (`last_lesson` = 2026-09)? ## What a segment is **Alex Tutor:** A **market segment** is a group of customers who need, buy and behave alike, so one marketing choice fits all of them. The dashboard's average blended several segments and hid every one. **Alex Tutor:** Before choosing columns, test the definition itself. Which of these is a segment? > **Exercise:** Which of these is a **market segment** in the sense of this lesson? **Alex Tutor:** You're cutting from **first-party data**: Klasna's own records of what people actually did. It beats guessing, but it only shows people who already came. **Alex Tutor:** Own data is the best start, but it has a blind spot. Where? > **Exercise:** Klasna's customer export shows the whole market, including families who considered Klasna and never signed up. **Sofia:** Which columns should I split by, then? **Alex Tutor:** Ones that change how people buy and behave: the goal, when they started, who pays, where they live. Not an ID. And never sensitive traits such as gender, ethnicity or religion: they're not in the file, on purpose. > **Exercise:** Which columns are good candidates for splitting the export into segments that behave differently? Select all that apply. ## What the average would have cost **Uliana:** So my six-lessons-a-month plan… **Alex Tutor:** …fits almost nobody. Spring crammers want more lessons, fast, and stop after the exam. Younger learners want fewer, for most of the year. Aiming at parents of 15-year-olds misses most of them, and the adults too. **Sofia:** So the plan needs groups, not a typical customer. **Alex Tutor:** Yes. Here plain rules on a few columns find them. When the groups aren't this obvious, clustering can suggest them; the Statistics for Product Managers and Marketers course covers that. > **Exercise:** What is the main risk of building the 2027 plan around "a 15-year-old who takes six lessons a month"? ## Your turn: tell Leo **Uliana:** Leo loves that dashboard. Can you explain, in two sentences, why it misleads? **Alex Tutor:** Use two numbers from the export. Don't attack the dashboard: the arithmetic is fine. It's the wrong summary for a plan. **Sofia:** Something he can read in ten seconds, then. **Alex Tutor:** Ten seconds, and a next step at the end: what you'll show him instead. > **Exercise:** Write two or three sentences for Leo explaining why the dashboard's "typical learner" (15 years old, about six lessons a month) is the wrong basis for the plan. Use at least two numbers from the export and end with what you'll show him instead. > **Key ideas** > > - An average can be correct and still describe almost nobody: no account in Klasna's sample is 15 and takes 5–7 lessons a month. > - Averages blend groups that behave differently. Here, spring exam crammers (40 accounts, 9.71 lessons a month, 30% back in September) hide next to younger learners at about 4 a month. > - A **market segment** is a group that needs, buys and behaves alike, so one marketing choice fits it. > - Segment from **first-party data** by what changes behavior: goal, start month, payer, country. Never by sensitive traits. **Alex Tutor:** Next: how to describe each group so you can find it, and which need actually wins it. - Previous lesson: [Case: the situation row](https://productovi.com/courses/marketing-strategy/the-situation-row-case.md) - Next lesson: [Find them by description, win them by need](https://productovi.com/courses/marketing-strategy/find-by-description-win-by-need.md) --- Productovi Learning. Free, no sign-up needed to read.