Notes — Bob Moesta on Jobs to Be Done

Four questions [Adler frame]

Q1 — What is it about? A methodology for understanding why people switch products: people “hire” products to make progress in their lives. The unit of analysis is not the product or the demographic but the struggling moment — the context+outcome vector that causes someone to act. Supply and demand are not as connected as assumed; demand is created by struggling moments, not by product launches.

Q2 — How is it argued? Through anecdote and case study: Snickers vs. Milky Way (demand-side competitive set); Southern New Hampshire University (60 anomalies → 200,000 students); Autobooks (three-demo buying-phase alignment → 4× conversion, half the sales cycle); Intercom (four distinct jobs → segmented pricing vs. HubSpot/Zendesk); condo downsizer (storage included → 30% sales lift); dining table condo (sacrificed 2nd bedroom → 22% sales lift). The method is qualitative and causal; Moesta explicitly rejects hypothesis-testing as the primary mode.

Q3 — Is it true? The core demand-side insight — that competitive alternatives are defined by the customer’s context, not by product category — is well-supported by the examples and aligns with positioning theory (Product Positioning). The four-forces model (push/pull/anxiety/habit) is a useful behavioural decomposition. Moesta’s claim that 10–12 interviews are sufficient (vs. statistical samples) is contested but is consistent with saturation logic in qualitative research. The critique that teams do JTBD “in a conference room” without actually talking to customers is correct and widely observed.

Q4 — What of it? JTBD reframes the roadmap from feature lists to struggling-moment queues — a more durable frame because struggling moments persist while technology changes. The method is actionable for PMs and founders: find 10 recent purchasers, extract the story, map the four forces. The hiring/firing lens also applies to employees (people hire companies; over 50% of job-switchers took no pay increase). Key design implication: reducing friction (F3+F4) can be more effective than adding pull (F2) — and adding features can actually increase anxiety.

Glossary

Job to be done (JTBD) — the progress a person is trying to make in a specific context. Not a task or a pain point; a vector defined by starting context + desired outcome. Products are “hired” to fulfil a job.

Struggling moment — the situational trigger that causes a person to perceive a need to change. Precedes all purchase decisions; exists independently of any product.

Push (F1) — contextual pressure that makes the current situation intolerable enough to consider changing (e.g. car has 280k miles, three large repair bills, long trip coming). Only pushes cause people to leave the old solution.

Pull (F2) — attraction to the anticipated outcome of the new solution. Ineffective without a push; more features increase pull but also increase anxiety.

Anxiety of the new (F3) — uncertainty about whether the new solution will work, fit in, or cause problems. Adding features can increase F3.

Habit of the present (F4) — inertia; the cost of giving up familiar workflows, relationships, and artefacts. Often underestimated (e.g. the dining table).

Four forces model — the behavioural system: switch occurs only when F1+F2 > F3+F4. Most product teams focus on F2; JTBD analysis addresses all four.

Six phases of buying — First thought → Passive looking (problem-aware, solution-unaware) → Active looking (both aware, comparing) → Deciding (trade-offs) → First use → Ongoing use (habit formation). Matching sales/onboarding to the buyer’s phase, not the seller’s process, drives conversion.

Demand side — the customer’s world: struggling moments, contexts, outcomes, trade-offs. JTBD starts here.

Supply side — the product/technology world: features, capabilities, business models. Traditional product development starts here and works backwards; Moesta argues this is backwards.

Hire and fire criteria — what caused someone to hire this product (and what would cause them to fire it). The operational expression of the job.

Clustering — grouping interview findings by shared pathways (push → pull combinations) rather than by demographic segment. Produces actionable job profiles; segmentation produces averages.

Three energy sources — functional (time, space, effort, knowledge), emotional (feelings, self-perception), social (how others perceive me). All three are present in any job; omitting social/emotional gives incomplete picture.

Bitchin’ ain’t switchin’ — complaints do not predict behaviour change. People will complain about missing features for years without churning. Only study people who actually switched.

Layers of language — pablum layer (surface pleasantries) → fantasy/nightmare layer (exaggerations) → what actually happened (the causal story). Interviews must get past the first two layers via contradiction and misparaphrase techniques.

Demand Side Sales — Moesta’s book; reframes selling as enabling the customer’s buying process rather than pushing the seller’s process.

The JTBD core thesis

“People hire products, they don’t buy them, they hire them to make progress in their life.”

The central reframe: competition is defined from the demand side, not the supply side. Snickers competes with protein drinks and Red Bull, not Milky Way. Milky Way competes with wine, brownies, and a run. Product-centric competitive benchmarks (calories, softness, ingredients) are irrelevant to the real competitive set.

Supply and demand are less connected than assumed. A struggling moment creates demand; a product launch does not. The struggling moment for online university (people who needed to go back to school but couldn’t attend in person) existed for years before Southern New Hampshire University found it in 60 anomalies and grew to 200,000 students.

Four forces model

The switch/no-switch decision is a system of four forces:

Switch happens only when F1+F2 > F3+F4.

Counterintuitive implication: adding features increases F2 (pull) but also increases F3 (anxiety — “can it really do all that?”). Reducing F3+F4 — removing friction, including storage, providing dining table space — is often more effective and cheaper than adding features.

Condo case: price raised, storage + moving included → F4 (habit of present: “what do I do with my stuff?”) eliminated → 30% sales increase. Dining table case: sacrificed 2nd bedroom for symbolic dining table space → F4 (habit of present: “where does the table go?”) eliminated → 22% sales increase.

Six phases of buying

1. First thought — a vague awareness that something needs to change.
2. Passive looking — problem-aware, solution-unaware; absorbing information without actively seeking.
3. Active looking — both problem and solution aware; comparing alternatives.
4. Deciding — trade-off resolution; making the final choice.
5. First use — onboarding; where anxiety of the new is highest.
6. Ongoing use — habit formation; the job is being done repeatedly.

Autobooks case: aligning sales demos to buyer phase (not seller’s close-focused process) halved the sales cycle and 4בd conversion. Three demos: (1) story/background for passive lookers, (2) alternatives overview for active lookers, (3) choice-framing for deciders.

Interview method

Who to interview: people who recently made the switch — recently purchased or recently churned. Never people who say they want to; “Bitchin’ ain’t switchin’.”

How many: 10–12 per job cluster. Patterns repeat around 7–8. Two rounds of 12 > one round of 24 (fresher perspective). Use design of experiments principles to sample within known market ranges; statistical significance is a proxy for knowledge, not knowledge itself.

What to extract: the story. Not opinions or stated preferences. Use misparaphrase technique (play back incorrectly → they correct and elaborate). Never ask a discussion guide of uniform questions — follow the path that has the most information. Get to the “edge of language” then bracket (“was it more X or more Y?” — neither answer is right, both prompt elaboration).

Layers of language: push past pablum and fantasy/nightmare to what actually happened. The causal story is always at the bottom.

Reference: Never Split the Difference (Chris Voss) — interrogation-as-therapy technique for eliciting stories.

Zero-to-one application: study what people are currently using and would fire when your product launches. For Facebook Marketplace: studied eBay, Etsy, Craigslist sellers and buyers. The jobs existed before the product.

Demand Side Sales and “help them buy”

Moesta’s book Demand Side Sales recasts the sales process from seller’s timeline to buyer’s timeline. The seller’s default is: qualify → demo → close. The buyer’s reality is: first thought → passive looking → active looking → deciding → first use → ongoing use.

Matching the process to the buyer’s phase eliminates friction from the purchase process itself. This is JTBD applied to the sales funnel rather than the product.

Parallel to Product Positioning (April Dunford’s “teaching the customer how to buy”) — both argue that the seller’s job is to reduce the buyer’s cognitive burden, not to push.

Clustering, not segmenting

Segmentation groups people by demographics → produces averages, not actionable jobs. Clustering groups people by shared push+pull pathways → produces job profiles that drive product, pricing, and positioning decisions.

Most products are hired to do 3–5 different jobs that are often in conflict (faster vs. more thorough). Identifying the conflicts enables deliberate decisions about which jobs to serve and which to decline.

Intercom case: four distinct jobs identified → four distinct pricing/feature bundles targeting different competitive sets (HubSpot for conversion, Zendesk for support).

What Jobs to Be Done is not

• Not pain-and-gain analysis (too supply-side; misses context).
• Not outcome-only analysis (misses starting point; value = distance travelled, not just endpoint).
• Not hypothesis-testing research (JTBD is hypothesis-building research; testing assumes you already know the questions).
• Not useful when there is no real choice (employer-provided health insurance, habitual categories like chewing gum where the “little hire” not the “big hire” is what matters).
• Not the same as Ulwick’s Outcome-Driven Innovation (ODI): Moesta/Christensen start from the person’s context; Ulwick starts from product functions. Only people have jobs; products don’t.

Choose what to suck at

“Choose what to suck at and figure out the trade-offs that you need to make and make sure that your trade-offs map the trade-offs of the customer.”

Most product failures stem from trade-offs the builder made that the customer didn’t agree with. QuickBooks: half the features, double the price — maps the trade-off of the target customer (doesn’t need more features, needs reliability and familiarity). Basecamp: adding Gantt charts and resource allocation would have destroyed the job it was hired for (simplicity).

Following your best/most-demanding users will destroy the jobs your lower-end users hired you for.