Crossing the chasm.

Most startups begin with a great idea and technical expertise. But there’s a profound difference between great technology and a great product. This post is for anyone who wants to build a product that people will actually use. Not just technology that solves a problem but won’t scale. The verb I use to describe this process is ‘productise.’ To productize means to turn something into a product.

I’ll explore the principles of productisation then drill down into a specific example. The example I use is x.ai’s Amy, the AI bot who acts as your personal assistant. Amy became de rigueur for scheduling meetings in the startup and VC community in 2016.

I chose Amy because she’s a great example of AI that’s been developed into a consumer product. Of technology that’s been successfully ‘productised.’

Amy’s also noteworthy because x.ai (the company who created her) profess to be ‘Anti-Lean’. As product people, we’re all familiar with Lean Startup. Lean has been the prevailing philosophy of product development for some time. So, it’s interesting to observe the emergence of startups proudly claiming to be Anti-Lean.

Thanks to x.ai.

If we imagine an axis of ‘lean-ness’, there are unicorns like Airbnb or Dropbox at one end. These are cast from the classic Y Combinator ‘lean’ mould. At the other extreme is a company like Tesla. No one could accuse Elon Musk of being a lean practitioner. Everything he does is a moonshot, requiring enormous capital investment.

Somewhere to the left of Tesla is a cohort of AI companies, like x.ai and Clarif.ai, who explicitly state, ‘Lean just doesn’t work for us.’ x.ai needed three years of R&D investment to get their product into the market. Anti-Lean requires more upfront capital. It also affords less opportunity for iteration and pivot. Anti-Lean companies need to be more certain about their users, and their users’ needs, when they do launch.

Why is productising important?

Productisation means translating a vision into a product that is usable and capable of global traction. The principal benefit of productisation is that it helps you cross the chasm. Geoffrey Moore famously identified a market chasm that separates companies who appeal to a small group of technology geeks from those who appeal to the mass market. His book of the same name explores all the factors that impacts a startup’s ability to make the leap. Productisation is one aspect of this.

Productisation also guards against Stack Fallacy. This idea comes from Anshu Sharma, once a Product Manager at Oracle and now a partner at Storm Ventures. Stack Fallacy suggests that, within any given technology stack, each provider overvalues the significance of their own layer. They also undervalue the significance of other layers, especially those higher up the stack.

Stack Fallacy is the reason “Database companies believe that SaaS apps are ‘just a database app’, which gives them false confidence that they can easily build, compete and win in this new market.”

According to Anshu, the companies who prevail are the ones that win at the user experience layer. They understand the end user’s context, motivations and desired outcomes best.

As a Product Manager, I support his view. In my experience, winning in the market means thinking about customer value, not individual screens. Productisation means focussing on outcomes rather than features.

We’re all familiar with the ‘jobs-to-be-done’ theory of products. People don’t use a product because they like the product itself. Instead, they rent them to achieve specific outcomes. Clay Christensen’s theory gained a lot of ground this year, I think, for good reason. I firmly believe that most utility products are tools that help people achieve outcomes. Only emotional aspects, like brand or aesthetics, really transcend it.

Productisation means focusing on the outcome the user is trying to achieve, rather than the underlying technology. This is more critical than ever thanks to exponentially increasing levels of competition. Two stats jumped out at me from Mary Meeker’s 2016 Internet Report. The first is that, in America, 80 per cent of people’s time is spent in just three apps. No prizes for guessing which apps. Nielsen data, published at the end of the year, indicates that the top 8 mobile apps are owned by just 2 companies: Google and Facebook.

The second stat is that 50% of people download zero new apps per month. This indicates the scale of the challenge new entrants face just to get attention. The adoption rate of new apps has plateaued because there’s too many to choose from. What this means for startups is, you only get a moment to make an impression. If you can’t communicate your value in that moment, you’ve lost.

The 3 Pillars of Product

What does it mean to be a ‘product’? There are three pillars: onboarding, core task completion and administration. You should think about your product in those terms. The vector that cuts across all three is user familiarity. To design each pillar, consider your users’ mindset as they interact with your product over time.

Users’ understanding of a product begins during onboarding. At this point, they’ll only have a minimal grasp of what you do. If they progress through onboarding, and begin using the product to complete tasks, they’ll inevitably become more familiar with it. If they continue to use it, and it becomes part of their lives, then administration of the product becomes increasingly important.

It’s possible to add a fourth pillar to the left of those three. The acquisition pillar. This is where the user first becomes aware of your product. I won’t cover acquisition in detail as it’s technically part of marketing, which sits outside of product. But it’s important to bear in mind as product and marketing form part of the same continuum.

Chamath Palihapitiya put it best when he observed,

“Users are only ever in three states — they’ve never heard about it; they’ve tried it; and they use it. What you’re managing is state change.”

Understanding the user mindset is key to the successful development of each pillar. You need to ask what the user is saying at each stage. At the acquisition stage, the user doesn’t know you, doesn’t know anything about you. What they’re saying is ‘Who are you and why would I use you? Why should I even care?’ During onboarding, they’re saying ‘I’m considering you. I’ve bothered to take time to play around with you. I’m thinking about you, but you still need to convince me.’

Once they reach core task completion, they’ve already passed a major milestone. At this point they’re saying, ‘I get you; I understand what you’re about and I can see the value.’ The administration pillar is all about customisation. The user’s saying ‘I get the value of what you’re providing, but I want to adapt you to my own needs.’

If you’re familiar with Dave McClure’s ‘pirate metrics’, it’s possible to map them against the four pillars. On the left are the acquisition metrics. Activation occurs during onboarding. Retention and revenue are core tasks. Referral is generally at the administration stage. Here, you can realistically expect users to advocate for your product.

Product Value

Winning at product means thinking about customer value. If you want to learn more about value within product, there’s a deck from Jerry Chen (from Greylock Partners where both Reid Hoffman and Josh Elman work) called Unit of Value that’s well worth reading. Jerry assesses the Unit of Value of unicorns like Salesforce, Dropbox or Docker. Then, he demonstrates how they deliver, and leverage that value across their service.

Products are ‘value delivery vehicles’. If you’re developing a new product, or have an idea for one, the fundamental question you need to answer is: what is the Unit of Value this will deliver? The answer will cut across how you price, scale and sell the product. Acquisition tactics communicate the promise of that value. Onboarding needs to demonstrate that value. The delivery of value occurs in the core task completion. Administration concerns the extension of value. It helps users get more value. It’s also where you can upsell additional products.

Once you understand your product’s Unit of Value, it’s easier to evolve it. Prioritising a long list of development requests is fundamental to the Product Manager role. I’ve found this 2×2 grid (known as the impact/effort grid) to be an effective tool for prioritisation within these vectors. Hunter Walk (who was Head of Product at YouTube) is a well known advocate, as are the team at Intercom.

To warrant consideration, all roadmap items should be in the upper half of the impact grid. If an item won’t help deliver your Unit of Value, it shouldn’t even be there. Strategic themes should be in the top right quadrant. In populating this grid with key stakeholders, Product Managers achieve a shared understanding of what the Product Team should, and should not, be doing.

By way of example, below is the 2017 roadmap for a product I worked on called Newsmart; a business English learning product that leverages content from The Wall Street Journal.

Any given product’s Version 1.0 will consist of just the task completion pillar. Nobody starts by building onboarding. This is just a route into the core product. To productise means to work outwards from this pillar and add the additional pillars. Productisation means establishing how people will find your product, understand what it is, use it, then customise it.

In the B2B space, there’s an additional level of complexity as buyers and users are different people. A buyer’s Unit of Value can be very different from that of the user. It’s easy to relentlessly improve the end-user experience and overlook the needs of the buyer, the person who’ll ultimately decide whether or not to invest in you. Products need to demonstrate and deliver value to each simultaneously.

Whereas the administration pillar is rarely given much thought in consumer products, in B2B it’s one of the most important elements. How does a buyer evaluate whether they’re getting return on their investment? Usually via a dashboard or via reports which tell them whether users are engaging with the product; and for how long.

With Newsmart, we overlooked the significance of this for far too long. We realised very late the importance of developing an administration suite that meant buyers could see value, not just users.

Products that start in the consumer space and pivot into the enterprise enter a very different realm. How you communicate and market changes. You don’t need to convince a single person anymore. You need to convince a variety of stakeholders. Succeeding in the enterprise means passing through a lot of barriers. Slack recently announced a partnership with Google’s G-Suite. That indicates, to me, that even they’ve been wrestling with this problem. They’ve got good traction with end-users, but are struggling to get their enterprise product into the market and recognise that Google are further down the track.

Product Design Patterns

If you examine the four pillars, the discussion around each is fairly advanced. Design patterns have been established that can be lifted and shifted into your own product. There’s no need to reinvent the wheel or go through the same learning curve each time.

As an example, let’s look at onboarding. Interesting characters exist, like Samuel Hulick. He’s established a whole business around this called User Onboard. His speciality is the ‘teardown’. Teardowns evaluate the onboarding of different products and break out the good and bad points of each. Product Managers can go to User Onboard and quickly understand best practice for different products across a number of verticals.

Samuel Hulick

Central to onboarding is the concept of the ‘Aha!’ moment. Aha! is when the user understands the value of a product and how it’s relevant to them. Succeeding at onboarding means getting users to the ‘Aha!’ moment as quickly as possible. As competition for user attention increases, time to Aha! needs to shorten.

Again, Chamath Palihapitiya is a great reference point. Chamath was head of the Growth Team at Facebook between 2008 and 2012. He’s very direct, often abrasive, in the way he communicates. This comes across in this classic talk on YouTube in which he rejects outright much of the pseudoscience about growth and ‘growth hacking’. For him, growth is all about understanding your product’s Unit of Value.

If anyone is sceptical about the importance of onboarding, I would cite Twitter as a great example of what happens if you don’t get it right. 2016 has not been a good year for Twitter. Business plateaued, growth fell off a cliff and its share price continued to flatline.

The heart of the problem is a failure to communicate their Unit of Value. Too many people still don’t understand Twitter, what it does, and why they should use it. That’s a failure of marketing and a failure of onboarding. A very high percentage of people register but never tweet or follow anyone. Failure to progress to these stages means you never reach the ‘Aha!’ moment. If you don’t reach Aha!, you don’t understand the value. If you don’t understand the value, you’ll churn.

To explore this further, let’s look at a successful example of onboarding. The reason I chose x.ai’s Amy is that she is fundamentally a piece of technology, not a product. Considerable resource was spent developing the AI and decisioning engine. When it moved into the market, it could have been taken in a number of directions.

x.ai made a number of choices that, I think, were brilliant from a product perspective. Amy’s onboarding is simple but effective. Once you’ve signed up, she emails you and says, ‘Hi, I’m your personal assistant, great to be on board,’ then asks you to take a couple of actions. All of us receive and respond to email requests. So, the interaction is familiar.

If you examine the first onboarding emails from Amy, she clearly states the value proposition. The ‘magic’, she says, ‘happens after 13 meetings.’ To demonstrate the value proposition, she schedules the first meeting with you. She sends an invite and asks, ‘Could you add a meeting reminder for this coming Thursday?’ By accepting the meeting, and seeing it drop into your calendar, you understand the value. There’s the ‘Aha!’ moment right there. All of a sudden, Amy has gone from being something novel and strange to something familiar. It’s a design pattern that everybody understands.

Amy could easily have remained a technology that schedules your meetings. In productising it, x.ai assessed our mental models and leveraged something known instead of introducing something alien.

Core task completion for Amy occurs in threads of emails between her and a variety of recipients. In these exchanges, she co-ordinates different calendars and, eventually, schedules a meeting.

In terms of administration, individual preferences for Amy are key. Most of us don’t want meetings at nine o’clock on a Sunday morning. I try and schedule most of my face-to-face meetings in the mornings, to leave the afternoons free, and my settings reflect this. Preferences change over time, of course. As settings are updated, Amy emails you to confirm.

For me, Amy is an excellent, contemporary example of productisation. The Unit of Value is time-saved. Amy calculates this based on your usage and sends updates which outline how many hours she’s saved you each month. This can amount to dozens of hours saved annually.

Payment

At some point, most products will take payment from users. This is another step that’s frequently overlooked. Design patterns for this are well developed too. A lot of research exists on ‘choice architecture‘ and the presentation of payment plans.

It’s also worth evaluating the extent to which your target users are able to pay. At Newsmart, we were predominantly selling into Asia.We launched with support for PayPal and credit cards via Stripe. Over time, a consistent message we heard from users was, ‘I want to pay, but I can’t!’ Our ignorance about preferred means of payment in China, Taiwan and Korea meant we were very slow to adopt certain methods.

If you’re Chinese, Alipay is probably you’re preferred method of online payment. UnionPay is also popular and PayPal is nowhere. Make sure you research the penetration of credit cards in your target markets. In many countries, outside the US, it’s very low — less than 50 per cent. If you’re not supporting preferred local payments, you’re leaving a lot of money on the table. Users might want to pay but can’t.

The Hook Model

As user psychology is essential to effective productisation, a design paradigm worth understanding is Nir Eyal’s ‘hooked model‘. A student of BJ Fogg, Nir draws on Fogg’s theories of behaviour. These regard behaviour as the outcome of motivation, ability and trigger. Nir has researched the common elements of ‘ultra-sticky’ products’. In his view, Facebook, Snapchat, WhatsApp, even LinkedIn, perfectly create the factors that effect behaviour change.

The stickiest products are the ones that nail the motivation and trigger moments. If the product is an app, the trigger will likely be a notification that says to the user, ‘Something is going on here, you’re missing out.’ Notice that notifications always pull you into the product. If you get an email from Facebook, there’ll be a call to action that takes you into the product. In Nir’s model, the most powerful motivator is known as ‘variable reward.’ Psychologically, this is more addictive than crack.

Variability injects an element of the unknown into the product experience. This generates a sense of FOMO that compels the user to take action. A combination of trigger and variable reward is what causes us to compulsively check Snapchat and Facebook. FOMO is also what prevents us from leaving the platform. How often have you heard someone declare they’re sick of Facebook and plan to stop using it? But no one ever does.

Productise — In summary

If you want to build a product, not just a piece of technology, start by thinking about value. Once you understand the value you provide, think about how you communicate, demonstrate and deliver it. Be mindful of the degree of familiarity users have with your product. And never lose sight of the outcomes they’re trying to achieve.