Product #115

Running Lean · Ash Maurya · 2010

Traction is evidence that people other than you and your team care about your idea. Traction is evidence of a working business model. Traction is the goal.

The business model is the product.

Start with problems before solutions. Customers don’t care about your solution, they care about their problems.

Prioritise testing what’s riskiest versus what’s easiest. Tackle your riskiest assumptions in stages. Answer ‘Should we build it?’ before you answer ‘can we build it?’.

You don’t need a working product to uncover problems worth solving - or even to land customers.

A minimum viable product (MVP) is the smallest solution that creates, delivers and captures customer value.

Most new products fail. Most startups that succeed (66%) had to drastically change their plans to do so. Starting with a great plan is less important than finding a plan that works before running out of resources. It’s important to have a process to quickly separate good ideas from bad ideas.

A Lean Canvas can replace a business plan and deconstruct your idea into clearly articulated assumptions.

The canvas doesn’t place emphasis on the solution and prompts you to think more broadly. Customers care about problems, Investors care about traction. You need to own the entire business model and make all the pieces fit.

The Business Model Design Playbook

1. Deconstruct your idea on a (Lean Canvas)
2. Test for desirability (Lean(er) Canvas)
3. Test for viability (Fermi Estimate)
4. Test for feasibility (Traction Roadmap)
5. Communicate your idea clearly and concisely (Business model pitch)

Tips for filling out the Lean Canvas

• Sketch the canvas end to end in 20 minutes, think of it as a broad brush stroke
• Complete as individuals first before discussing as a group
• You can leave boxes blank and come back to them
• Be concise - respect the constraints of the single page, brevity can provide clarity
• The time horizon should be immediate - what are you going to do next?
• You can complete in any order. Loop back on yourself to make sure things stick together
• Don’t confuse customers with users

Your unique value proposition should answer: Why is your product different and worth paying attention to? Distill the essence of your product into a few words that could be the headline on your landing page.

• Focus on outcomes. Benefits over features:
• Feature: professionally designed templates
• Benefit: eye-catching resume that stands out
• Outcome: landing your dream job
• Answer what, why and who

When you try to reach everyone you reach nobody. Identify your first customers.

Price is part of the product, it defines your customers. Getting paid is a big part of idea validation.

Use outcome metrics versus output metrics (MRR, LTV, New Customers). Use leading metrics where you can - so you can see how things are doing in real-time.

Stress Test Your Idea for Desirability

The best way to get a customers’ attention is by leading with your unique value proposition. A compelling UVP promises a better desired outcome, a better way of achieving the desired outcome or both. Focus on who you’re targeting and the problems that stop them from achieving their desired outcomes.

The innovators gift: new problems come from old solutions. Frame the problem as existing obstacles with old solutions that prevent customers from achieving their desired outcomes.

Once upon a time, there was a [customer]. Whenever they needed to get a certain [job] done, they would pick [existing alternative]. One day, that existing alternative broke because of [switching trigger]. Because of that, the [customer] realised that the [existing alternative] wasn’t the best choice for the [job] because of these [problems]. This realisation pushed the [customer] to search for a better solution and consider other alternatives. Until finally they found a [new solution] that helped them get the [job] done better.

Cause a switch by anchoring new solutions against problems caused by old solutions.

A switching trigger is when we realise our existing solution is no longer good enough to get the job done.

The Customer Focus Model describes the casual forces (Push, Pull, Inertia and Friction) that shape how people select and use a solution for a specific job-to-be-done:

• A switching trigger provides a motivation to switch (a bad experience, change in circumstance, or an awareness event)
• Inertia and resistance to change means new solutions need to be much better (3-10x better) than the status quo to overcome the friction associated with switching.
• A switch starts when attractor forces are greater than the detractor forces: Push + Pull > Inertia + Friction

Stress Test Your Idea for Viability

The trap of financial projections is focusing on the outputs not the inputs. Its the inputs to the business model that matter.

Don’t create a financial forecast → use a fermi estimate instead. A bottom-up approach, starting with a set of inputs, roughly estimated with assumptions. Then test the assumptions. We can stress test viability with just 5-7 key metrics. Traction is the goal. Traction is the rate at which a business model captures monetisable value from its customers. Monetisable value doesn’t have to be revenue, but it’s a future indicator of revenue

Testing viability with a Fermi estimate:

1. Define a minimum success criteria MSC. The smallest outcome that would make you deem your project a success. If a startup frame it as ARR(Annual Recurring Revenue).
2. Assess if your customer funnel can deliver on this goal? Input your best-guess estimates for the pirate metrics (AARRR). Estimate the active customers you’ll need, your customer acquisition rate, churn, customer lifetime value, estimate the number of leads, estimate your referral loop.
• If you have an MSC goal of achieving $10m ARR in three years to achieve that goal you’ll need to acquire one of the following:
• 1 million customers paying you $10/yr
• 100,000 customers paying you $100/yr
• 10,000 customers paying you $1,000/yr
• 1,000 customers paying you $10,000/yr
• 100 customers paying you $100,000/yr
• 10 customers paying you $1,000,000/yr
3. If things don’t look promising, redefine your model. It’s better to invalidate your model in 5 minutes than to spend 5 months pursuing it. A Fermi estimate also helps identify the riskiest assumptions.

Stress Test Your Idea for Feasibility

Don’t create a product roadmap - use a traction roadmap instead. Product roadmaps assume you know what you’re building traction roadmaps don’t.

If you set your success criteria (e.g. how many active customers do you want 3 years out) you can work backwards by assuming different growth rates to demonstrate how many customes you’ll need at each point. Note lower growth rates require more customers earlier / higher customer acquisition rate at the beginning.

Formulate a Now-Next-Later rollout plan. Avoid premature optimisation - prioritising the wrong things at the wrong time. At any given point there are only a few key actions that can have the biggest impact on your business model:

• Now: Problem/Solution fit (3-6 months):
• Uncover customer needs
• Identify what your MVP needs
• Secure tangible commitments
• You don’t need a working product to acquire paying customers
• Aim for an evidence-based go/no go decision to move forward to the build stage.
• Next: Product/Market fit (18-24 months):
• Build the first iterations
• Demonstrate value delivery
• Improve product continuously
• You don’t need lots of users at this stage
• Later: Scale
• Pursue growth
• You only need to focus on one engine of growth at a time

Communicate Your Idea Clearly and Concisely

An elevator pitch is a quick overview of your idea that you can present in 30 seconds. Most fall flat because they assume too much and are too solution centric. Instead focus on why your product needs to exist.

Part 2: Validation

The Continuous Innovation Framework: Addressing your weakest link is the only thing that matters, everything else is premature optimisation. Identify problems → generate a diverse set of solution candidates → make bets on the most promising ones → test → decide on next actions.

Validate Your Idea Using 90-Day Cycles

Baseline your system, find the limiting constraint and get to the root cause behind the limiting constraint. This should expose the riskiest assumptions of your business model. Once you’ve moved the constraint start the cycle again.

Generative or discovery experiments uncover key insights that help you formulate better hypotheses, which you then verify through evaluative or traction experiments.

• Seven habits for highly effective experiments:
1. Declare your expected outcomes upfront
2. Make declaring outcomes a team sport (first as individuals, then as a team)
3. Emphasise estimation, not precision
• Search for analogs
• Use your traction roadmap and AARRR model
• Start with ranges instead of absolutes
4. Measure actions not words
5. Turn assumptions into falsifiable hypothesis
6. Time-box experiments
7. Always use a control group

Problem/Solution Fit Playbook:

• Before building your product you should demonstrate there is sufficient demand for it.
• Use a Demo-Sell-Build process to create customers without creating a product.
• Use smoke tests, landing pages, webinars, preorders, crowdfunding, direct sales or mafia offers to to validate demand.
• You’re done with problem/solution fit when you can make an evidence-based go or no-go decisions on whether to move forward with your business model to stage 2. You can:
• Repeatedly attract, activate, and cause a switch with your early adopters (Desirability).
• Secure sufficient tangible commitments (e.g., advance payments, letters of intent) from early adopters, as defined by your traction roadmap (Viability).
• Clearly define the smallest thing you need to build (your MVP) to deliver value to your early adopters.
• A mafia offer is an offer your customers cannot refuse because it solves their top problem(s) and provides a clear way to get started. Steps to build a mafia offer:
1. Understand the problem through customer interviews.
2. Define the solution that directly addresses the problem.
3. Test the offer through campaigns before building the MVP.
4. An offer includes a unique value proposition, a demo, and a call-to-action (CTA).
• You don’t need a working product to start selling; instead, focus on selling the promise.
• Conduct face-to-face interviews to uncover the real pain points and problems of customers.

Use the Customer Forces Canvas to capture insights from discovery sprints:

• PUSH (motivation for progress): Identify what changed in the interviewees’ environment that pushed them to get a job done:
• Triggering Event: the thoughts and events that moved them from passively looking, to actively looking for a solution.
• Switching Trigger: The causal event that caused them to switch. Typically a bad experience, change in circumstances or an awareness event.
• Desired Outcome: What’s the desired outcome at the start? How would they measure success?
• What’s at Stake: What were the consequences (if any) for ignoring the triggering event(s) and doing nothing?
• PULL (attraction of the chosen solution): what’s attracted them to their chosen solution:
• Consideration Set: What other existing alternatives were considered for the job?
• Chosen Solution: The chosen solution that was hired for the job and how they found it
• Unique Value Proposition (Promised Outcome): the specific appeal of the chosen solution? Why did they pick it over alternatives?
• Expected Outcome: What did they expect to achieve with the chosen solution? What specific metric(s) would they use to measure success?
• INERTIA (resistance to changing the status quo)
• Existing Solution: how are they currently getting the job done?
• What’s Broken: Problems with existing solution as a result of the switching trigger?
• Barriers to Switching: Any existing habits or switching costs
• FRICTION (resistance to using the chosen solution)
• Anxieties: fears or worries the person expressed as they started using their chosen solution.
• Barriers to Adoption: challenges the person encountered with the chosen solution during usage.
• Additional Details: any additional insights about the chosen solution.
• WHAT’S NEXT
• Actual Outcome: What was the actual outcome after using the chosen solution?
• Next Summit: What’s the next action for this person? Was the job done well enough? Will they continue using their chosen solution or consider using a new solution?

Design Your Solution to Cause a Switch

A product that causes a switch offers customers a better way to achieve their goals without the existing problems.

The MVP must significantly outperform existing alternatives, with a 3x–10x improvement needed to provide a strong incentive to switch.

Identifying the primary struggle in the customer journey is the first step in designing an MVP. Primary struggles often arise from dissatisfaction, friction during usage, or friction during solution selection.

Crafting a compelling promise involves shifting focus to what is different and better about your product. Identify key attributes to improve and map them to your product’s strengths, such as speed, performance, or simplicity. It’s important to align with customer values and not just focus on functionality.

Defining switching triggers is key to targeting the right customers; they could be experiencing bad outcomes or looking for a better way.

Consider the three common MVP approaches: Concierge MVP, Wizard-of-Oz MVP, and Foot-in-the-Door MVP.

Ensure that your MVP addresses a specific problem, delivers a compelling promise, is priced fairly, targets the right people, and is packaged for quick deployment.

Deliver a Mafia Offer Your Customers Cannot Refuse

A Mafia Offer is an offer your customers can’t refuse. The goal of the mafia offer is to secure enough tangible commitments from early adopters to warrant building your MVP.

A good pitch acknowledges popular existing alternatives (the competition) and demonstrates how your solution is better. A strong narrative will help your pitch. Consider using a storytelling template like the Hero’s Journey.

Practice, active listening, constant testing and iteration will result in a great pitch.

Be prepared to run 2 sprints over 4 weeks and pitch to 20-30 people. You should get to a 60-80% conversion rate from qualified leads to your customer specific call to action.

3 steps to run an offer delivery sprint: Assemble, deliver and optimise your offer.

Use your existing components: triggering event, the stakes, the problems, your unique value proposition then end with a strong call to action.

Demo your new, better way. Showcase your UVP with the smallest thing possible in order to maximise the speed of learning.

Position your MVP as a prize (use ‘early access’ not Beta). Charge from day one (include your pricing model in the call to action). Make a logical case for how you arrived at a fair price anchored against existing alternatives and the value you deliver.

Look for the key constraint, and look for ways to improve it.

90-day Cycle Reviews

• Collect and update your artefacts: elevator pitch, lean canvas, traction roadmap.
• Assemble a progress report pitch deck. Include the following slides:
1. Review of the 90-day cycle objectives
2. Elevator pitch
3. Lean Canvas snapshot
4. Traction roadmap snapshot

What we did:

5. Validation campaigns you selected
6. Experiments you ran

What you learned:

7. Insights and key learnings
8. Traction: summarise the results from your campaign

What’s next:

9. Current constraint in your business model
10. 3P Next Action: Persevere, pivot or pause.
• Invite your key stakeholders
• Deliver your progress update in 10 minutes (20% of the meeting) leave 80% of the meeting for feedback and decisions.
• Solicit advice. Avoid success theatre.

Part 3: Growth

Get Ready To Launch.

Once you have your first couple of customers, it’s tempting to shift your focus from acquisition to product development (adding features). Instead get just enough traffic to support learning, then focus on key constraints - many of which are likely to be in the acquisition funnel (e.g. replacing high-touch interactions with automated touchpoints).

Be razor focused on getting to MVP release. Set a nonnegotiable launch date and fight scope creep.

Create a company wide dashboard so you can assess how your business model / funnel is performing (AARRR). Use fewer metrics to help you focus on the right things. Strive for actionable metrics. Use cohorts to measure and visualise the effectiveness of your business model / funnel over time - by pitching one batch of users against another. Cohorts can help you home in on causality - by inspecting what changed for that batch

Make Happy customers

Focus on helping customers achieve their desired outcomes. Once you’ve got some level of repeatable acquisition focus on activation. Acquisition is where happy customers are made - get them to the ‘Aha moment’. Creating happy customers improves retention, revenue and referral.

If you have 5 sprints, focus 2 on getting customers to the aha moment, and 3 on getting to their switching moment (retained them enough times to go all in on your solution).

Immediately after launch 80% of your time should be spent measuring and improving existing features, rather than chasing new shiny features.

Outlearn the competition: speed of learning is the new unfair advantage.

Trigger your customers: set the right expectations, use reminders, share best practice, prompt the next JTBD once the previous one is completed. Establish regular touchpoints (weekly activity reports). Nudge with email. Integrate your product into their existing routines.

Help your customers make progress: reduce the paradox of choice, allow them to experiment and provide high-touch support. Continuously improve your product. Share customer case studies. Make giving feedback easy.

Reinforce progress: show progress indicators, celebrate customer wins, give meaningful gifts.

Find your Growth Rocket

• Once your customer segments demonstrate regular use of the product and are making progress toward their desired outcomes → it’s time to focus on growth.
• In the growth phase you have to move from high-touch to more scalable paths to customers.
• The Rocket Ship Growth Model
• Three key parts:
• Payload: your core product
• One or more booster rockets: non-scalable customer acquisition channels
• Spacecraft: your primary scalable customer acquisition channel
• Each part has it’s own fuel (time, money, content, users) and helps you gain altitude (traction). The range of the rocket is determined by the efficiency of its engine and fuel.

Growth rockers utilise a flywheel (growth loop) in their engine design that regenerates propellant, which drives sustainable growth (traction).

• Three stage process:
• Define your mission. Sketch out how many boosters you need (non-scalable channels) and how you’ll power your spacecraft (scalable channel).
• Validate your design assumptions before lifting off (desirability, viability, feasibility)
• Identify, test and utilise non-scalable booster rockets (PR, direct sales, events)
• Optimise your core happy customer loop.
• Introduce additional boosters as needed until you achieve product/market fit
• Fire your growth rocket - your sustainable flywheel or growth loop
• You can attract new customers from the actions of past customers if you reinvest part of the value you capture into customer acquisition:
• E.g. Money, Content, Data, Referrals, Retention
• If using money make sure LTV (Customer Lifetime Value) is higher than CAC (Cost of customer acquisition).
• Building a growth rocket requires a renewable propellant and an efficient engine.

Full Book Summary · Amazon

Subscribe Button 

Quick Links

A quick product simplicity test: remove all the explainers · Article

Will AI outsmart human intelligence? LLMs are immortal · Video

Is your product at risk of Product Market Fit Collapse? · Article

Taste is upheld through: authority, systems and energy · Article

About empty states for prompt fields · Article

Let’s Verify Step by Step

Hunter Lightman. Vineet Kosaraju. Yura Burda. Harri Edwards. Bowen Baker. Teddy Lee. Jan Leike. John Schulman. Ilya Sutskever. 2023. (View Paper → )

In recent years, large language models have greatly improved in their ability to perform complex multi-step reasoning. However, even state-of-the-art models still regularly produce logical mistakes. To train more reliable models, we can turn either to outcome supervision, which provides feedback for a final result, or process supervision, which provides feedback for each intermediate reasoning step. Given the importance of training reliable models, and given the high cost of human feedback, it is important to carefully compare both methods. Recent work has already begun this comparison, but many questions still remain.

We conduct our own investigation, finding that process supervision significantly outperforms outcome supervision for training models to solve problems from the challenging MATH dataset. Our process-supervised model solves 78% of problems from a representative subset of the MATH test set. Additionally, we show that active learning significantly improves the efficacy of process supervision. To support related research, we also release PRM800K, the complete dataset of 800,000 step-level human feedback labels used to train our best reward model.

From using the latest Open AI models it’s clear process supervision is creating models that can answer more complex queries and gets us one step close toward agents. The PRM800K dataset is going to make LLMs so much better at math problems.

Book Highlights

When YouTube leadership set the one-billion-hour daily watch time goal, most of our people judged it impossible. They thought it would break the internet! But it seemed to me that such a clear and measurable objective would energize people, and I cheered them on. John Doeer · Measure What Matters

Most of the mistakes I’ve made, seen, or heard about didn’t happen from selecting the wrong tool or method. Instead, most of the mistakes came from the following situations: Prototyping either too much or too little. Prototyping the wrong thing. Not setting expectations for what the prototype will be. Todd Zaki Warfel · Prototyping

The product manager is responsible for defining-in detail-the product to be built, and validating that product with real customers and users. The product marketing person is responsible for telling the world about that product, including positioning, messaging and pricing, managing the product launch, providing tools for the sales channel to market and sell the product, and for leading key programs such as online marketing and influencer marketing programs. Marty Cagan · Inspired

Quotes & Tweets

A brand is not what you say it is, it’s what they say it is. Marty Neumeier

All models are wrong but some are useful. George Box