I covered the Customer Discovery flow for my web startup in Part 1. Here I’ll be covering the next step: Customer Validation.
At the end of Customer Discovery, you should have identified a customer problem worth solving and started building your solution (MVP). During Customer Validation, you’ll test your finished MVP by selling it to earlyvangelists and, in the process, start developing a repeatable and scalable sales process.
While Customer Discovery was all about Problem/Solution fit,
Customer Validation is all about Product/Market fit.
The good news is that since web apps are primarily distributed through a website, developing a repeatable and scalable sales process (conversion funnel) is easier for a web startup than, say, for Enterprise software, which usually requires multiple stakeholders, necessitate face-to-face selling, and eventually a sales force to scale. The bad news is that relying on just a website to sell is much harder — You only have 5–8 seconds to make an impression, and it’s harder to troubleshoot the sales process without face-to-face interaction with customers.
I’ve found that “Getting out of the Building” is just as important during Customer Validation as it was during Customer Discovery.
Customer Validation: Have I built something people want?
Here’s my Customer Validation Flow (you’ll probably want to click to enlarge and skim it before reading on).
The 3,000 Foot View
Before you’re ready to sell, you have to distill down your product into a clear message (positioning), develop your sales materials (demo, website), identify your preliminary distribution channel, define your sales roadmap (conversion funnel), and of course, finish your MVP. You then test your finished MVP + sales process by first selling face-to-face to earlyvangelists, then to web visitors. Product/Market fit is all that matters here, so build/measure/test until you’re fit.
Let’s Get Ready to Sell
Articulate a Unique Value Proposition
Your Unique Value Proposition is a single, clear, compelling message that states why you are different and worth buying. This is what customers see first and is arguably the most important element on your landing page. A good starting point for crafting a compelling UVP is revisiting your prioritized list of problems from Customer Discovery and answering “what, who, and why.”
Here’s an example of the current UVP I am using for CloudFire:
Photo and Video Sharing for Busy Parents.
No uploading. No reorganizing. No hassle.
What is it? A photo and video-sharing service
Who is it for? Busy Parents
Why is it different from what I use today? It’s hassle-free: You don’t have to upload or reorganize your photos and videos.
For more on positioning, read/reread Al Ries/Jack Trout’s classic book — Positioning: The Battle for Your Mind. Crafting a good UVP is hard work, but don’t worry about perfecting it. Like all your earlier hypotheses, this one is a guess, too, that you’ll be testing and refining later.
Build a Product Website
With your UVP crafted, it’s time to polish up your MVP demo (from Customer Discovery) and start building your product website. While there are many ways to structure a product website, this is what I feel are the minimum set of pages to include:
1. Landing Page: State UVP, link to a demo, strong call-to-action
While there are other important elements like social proof and credibility that eventually need to go on your landing page, you may not have all of these on day 1, and the UVP is the most important element to test first anyway.
2. Pricing Page: How much does the service cost?
This is where you detail your pricing. Several tactics to lower buyer friction include offering free trial periods, money-back guarantees, deferred credit card payments, etc. Pricing is more art than science, and you’ll need to test what works best for your customers. I will, however, restate my position on NOT giving away the product for free. Anyone can give away a product. In return, you don’t learn anything about your customer’s willingness to pay.
There is a line of reasoning that suggests implementing a business model after product/market fit so as not to add friction. I believe that strategy is sound only if you built your product absent Customer Discovery and don’t know if you have Problem/Solution fit. The point of Customer Discovery was finding a problem worth solving which involved determining the price a customer would pay for a solution to that problem. Your job now is to test if that solution could be your solution. Lowering buyer friction through free trials and/or money-back guarantees is all good and expected, but now is not the time to back down on pricing to make a sale — as appealing as it may seem. Even if you’re considering a Freemium model, I would not offer a free plan till after Product/Market fit. You must validate that your customers will pay for your premium plan first.
The exception is if you are building a service that *primarily* relies on high-network effects to succeed. For example, social media services (Twitter/Facebook) and marketplaces. There, early user engagement trumps revenue.
3. Sign up Page: Activate users
It goes without saying that the sign-up process needs to be as painless as possible to get customers activated and use your product as quickly as possible. Minimizing steps, only asking for what you need, deferring registration to later or never (lazy registration), connecting with Facebook, etc. are all ways to reduce signup friction.
4. About Us: Company positioning
People want to know who you are and what qualifies you to offer this service. For inspiration, read Jason Cohen’s: You’re a little company, now act like one.
5. Tour: Features/Benefits, How it works?
Not everyone will view the demo. Our usability testing showed many users abandoning a 2-min demo after 45 seconds. Some wouldn’t even watch the video for fear of getting viruses on their computer (yes, probably not our target demographic). Having a separate tour page also allows you to expand upon the other top problems you solve that may not be covered in your UVP.
Define Conversion Funnel
With your website pages created, you can now string them into a coherent sales roadmap or conversion funnel. Your conversion funnel should chart your web visitors from when they hit your landing page to the point when they trigger a revenue transaction.
Dave McClure has captured the essence of the conversion funnel with his AARRR startup metrics which has five basic steps:
- Acquisition: How do users find you?
- Activation: Do users have a great first experience with your product?
- Retention: Do users come back and use your product?
- Referral: Do users tell others about your product?
- Revenue: How do you make money?
Using the pages above, a typical conversion funnel would look like this:
Landing Page -> Pricing Page -> Signup Flow -> Invite/Tell a friend -> Upgrade Account
While optimizing steps (like combining the landing page with pricing/signup) is possible, at least during Customer Validation, I prefer to organize my conversion funnel for maximum learning versus maximum conversion. For instance, by keeping my landing page completely free of pricing, I can measure the effectiveness of the UVP without having to guess if price was a factor.
Select Customer Acquisition Channels
Customer acquisition channels for a web-startup range from free channels (blogs, SEO, referrals, etc.) to paid channels (SEM, paper ads, partners, etc.). The key objective during Customer Validation is NOT growing customer acquisition but validating product/market fit. You need to drive “just enough” traffic to learn and optimize for product/market fit. You can usually accomplish this solely by relying on free channels, but it is okay to supplement with some paid channels (like SEM). However, be wary not to overspend at this stage which is a common tendency.
I jumped into paid advertising (the expensive kind) too early with my last product BoxCloud. While it drove lots of good traffic, they all ran into the same product/usability roadblocks. I would have learned the same lessons with one-fifth the traffic and one-tenth the cost.
That said, free channels aren’t free, either. They take time to develop. Now is the time to invest in these (before you need them). Publishing your website publicly (versus hiding it behind a closed beta) is already a good start, as you’ll start gathering some SEO value. Another channel worth developing is starting a blog. But don’t make the mistake of simply touting your solution or rambling on about your product roadmap or release schedule. Nobody cares (not yet).
Customers care about their problems NOT your solution.
- Dave McClure (paraphrased)
Instead, blog about the problem space you just validated in Customer Discovery. As Gary Vaynerchuck once said: “If you want to launch your brownie, become an expert in brownies first.” Earn trust. Be passionate.
This blog drives lots of traffic to my products — albeit indirectly. I started a more direct CloudFire blog here that needs more love.
Get MVP to “It Works”
The final step to getting ready to sell is getting your MVP to a state where Earlyvangelists can use it. There is a common misconception that an MVP is a quick-and-dirty release engineered to solicit early feedback from customers. While speed is a key consideration in a lean startup, so is quality. The finished MVP needs to be a well-executed implementation of the minimum feature set identified in Customer Discovery, and it has to work.
Sell to Earlyvangelists
With the finished MVP, it’s time to visit with the original interviewees from Customer Discovery and sell them your MVP. In addition, I ensure to include several new potential customers that haven’t been exposed to any product discussions. The reason for this is to test your sales process as closely as possible to how a web visitor would but with the added benefit of being in the room.
This is how I structure my sales presentation — part usability testing, part selling:
1. Show customers your landing page.
2. Ask them what they think the service is about. What’s compelling, and what’s different? Does the difference matter?
3. Show them the pricing page and test pricing.
4. Are they convinced to try the service? If not, why not?
5. If they are sold, watch them go through your sign-up flow.
6. Take lots of notes and fix obvious usability/messaging issues before the next iteration.
Structuring the sales presentation like a usability test reveals lots of actionable problems early that you can address before going on to testing web visitors. To learn more about running interviews in a usability test format, read Steve Krug’s latest book: “Rocket Surgery Made Easy.”
If you’re going to be relying on free customer acquisition channels, now is the time to assess their effectiveness in driving new web visitors to your landing page. Otherwise, supplement with some paid traffic.
Measure Product/Market Fit
Now that you have some web traffic, you need to instrument a basic conversion dashboard that measures your AARRR metrics. Of the five, Activation and Retention are the most critical for achieving Product/Market Fit. Focus on those first, and don’t get distracted by trying to grow customer acquisition or referrals (virality). You’ll optimize these in the next stage — Customer Creation.
Getting to fit typically requires multiple iterations on the UVP (positioning) and initial user experience flow. Because initial web traffic tends to be small at this stage, I have found it more effective to balance quantitative testing (A/B tests) with more qualitative testing (face-to-face usability tests). Usability experts agree that you need just three face-to-face tests to reveal 80% of the issues. Try learning anything from 3 abandoned web visitors.
Quantitative metrics clearly have the advantage of scale once you’re generating lots of traffic, but they also have a place early on, especially for troubleshooting problems with your conversion funnel. For instance, I’ve used my activation metrics to uncover several technical issues with my signup flow that were browser/OS specific and prevented customers from completing the installation of CloudFire.
Iterate or Exit
So how do you know when you’ve hit Product/Market fit? Dave McClure suggests hitting a six or higher on a customer satisfaction scale of 1–10. Sean Ellis uses a survey to determine if more than 40% of his customers would be disappointed if the service were discontinued. The mechanics of measuring Product/Market fit is a little vague and subjective — how many people do you poll, when, and how?
I rely on a combination of my revenue and retention metrics to identify customers that are continually paying AND using the product. I then poll these customers with the “would you be disappointed” question. I feel that question is less vague than relying on a 1–10 scale whose weighting might vary from person to person. As with Problem/Solution fit, there is no empirical answer to how many customers you need to declare a fit. You stop when you aren’t learning anything new. Getting a strong signal from just 30 customers could be enough to declare Product/Market fit and move you to the next stage — Customer Creation.
Update: The workflow described in this post has been refined even further and turned into a book: Running Lean — with step-by-step guides, techniques for finding prospects, and field-tested interview scripts.
You can learn more here: Get Running Lean.