Discuss Home · Bootstrapped Podcast · Scribbleton Personal Wiki · HelpSpot Customer Service Software · Thermostat NPS

Build or validate?


#1

Hi there,

I have a business idea. After one week researching about alternatives and possible competitors, I think it has potential. It also fits the type of product/business I want to build.

The product I’m thinking about is a mix between Charlie and Rapportive, targeted to Shopify store owners.

Now I’m trying to decide what should be my next step. I’m considering two options:

  1. Build the simplest possible (polished) version and launch it.
  2. Try to validate the idea with potential customers.

Build basically means:

  • Partnering with a coder (I’m a product manager).
  • Spending some money.
  • Knowing for sure, once I’ve launched, if people wants it or not.

Validating the idea should be less risky, but I’m not sure how to do it without getting false positives or negatives.

My app requires accessing the emails and customer database of Shopify store owners, so it requires from customers a good amount of trust, which I’m not sure is compatible with offering the service in “concierge mode”.

What would you do?


#2

I wouldn’t spend a penny or write a line of code until you talk to 20 Shopify store owners directly, explain the service, and find out how many of them would pay for it.


#3

Thanks for your feedback @coreysnipes.

How do you know they will actually pay, even if they say they will?

Or, how do you know the wouldn’t pay, if they can’t test the product by themselves in a real scenario?


#4

Cold sales aren’t easy. If you can at least get a few replies saying people are interested that can improve your confidence in the product selling.


#5

A good time to share Amy Hoy’s link:

http://unicornfree.com/2014/validation-is-backwards


#6

Thanks for the link @shantnu. Do you have experience with this type of process?


#7

How do you know they will actually pay, even if they say they will?

You don’t. But you’ll have a much better feel for the “real” answer at the end of the process. You’ll also have a couple of things…

  • A much better understanding of 20 real customers’ perceived needs
  • A clearer idea of which of the features you’re considering would be more or less important
  • A much more polished pitch that explains the problem/solution in your customers’ terms

Plus (arguably most important) you’ll have some basis on which to decide whether to move forward (or not). Right now, all you have is your assumptions about the market & some competitor research.


#8

@CescVilanova
Here is a link to one of the best articles I have ever read on this subject. You should try out this out, to validate the product before you execute. It’s simply genius.

(Basically trick /make people think the product is live, see where they click (what plan etc.) - and then the 2nd page (oups! you caught ‘the product’ too early). Personally, I am not a believer in asking (surveying) whether they would use/buy it or not. Unless you really try to cold sell it and get money on your account before building it.
http://blog.bufferapp.com/idea-to-paying-customers-in-7-weeks-how-we-did-it

Buy a subscription at Squarespace / Wix or whatever (if you you’re not HTML / CSS savvy.) and build a landing page! Looking forward to hear what you decide to do :slight_smile:

Good luck.


#9

I’ve gone both the validation route and the “sales safari” (Amy Hoy) route…the latter is the only time I’ve been able to get paying customers. In all fairness, maybe it was how (badly) I did customer interviews/research,
but a good/helpful book is The Mom Test by Robert Fitzpatrick. Unfortunately I found this after failing to get paying customers and spending time/money to build something that people said they’d pay for but never did.

The buffer example is great, as well as what Josh Pigford has done with Baremetrics:
https://baremetrics.io/blog/idea-to-5000-in-5-months He also did an AMA on Reddit (http://www.reddit.com/r/Entrepreneur/comments/2kf2xt/i_bootstrapped_a_saas_business_from_idea_to/)

For me, it’s been more tangible to find a problem that people are actively complaining about/seeking a solution, vs. trying to find people that have the problem you intend to solve, then confirming your solution and then getting them to pay. More often than not, people don’t realize they have the problem and to pull it out of them doesn’t make it easier to get them to pay for it’s solution. Of course in talking to people you may stumble across another pattern of problems that people do want solved that is so far from your original idea, but that could be profitable. Good luck!


#10

I am not sure I like the idea of tricking potential customers into thinking a product exist that @Reshal mentioned. Why not just a landing page with an email collection form offering to sign up for a private Beta?

Here is one validation strategy that can also be used for getting early customers:

  1. Create a twitter account with a description of the product in the profile of the account
  2. Link Twitter profile to a web site with a single landing page with more info about your product and email collection form. This is a GREAT article on building effective landing pages.
  3. Start tweeting about the topic that your prospective customers want.
  4. Search for accounts to follow using keywords you think your customers would have in their profile. You should get 5-10% of accounts to follow you back, depending on how well you can locate/define your potential customers. I suggest using ManageFlitter for search.
  5. Track how many follows and/or emails you get and optimize search for more accounts weekly.
  6. When you build up a good list, send an email and ask some questions.

Good Luck!


#11

@Webbie - maybe an incorrect choice of wording.
Your prescribed Twitter strategy to validate is very effective if you are able to target your exact audience. Which ManageFlitter certainly helps with doing. However I think you could combine to get benefits of both worlds.

  1. Find your audience through Twitter (if you don’t already have a big network of potential customers) using tools such as ManageFlitter.

I see two options, and you could even experiment by trying out both.
2a) Create a landing page (without showing that it hasn’t launched - or beta) - see if they click further on, maybe even choose a plan? (determine if people are interested in this product and are continuing to the perceived payment page, lastly make them sign up - as Joel from Bufferapp.com did.)

2b) Create a ‘coming soon’ landing page. And ask people to sign up.


#12

I’m not sure the validation route and the sales safari route should be opposed. I know Amy feels strongly against validation but in my opinion she describes a wrong type of validation. Of course validation doesn’t mean you expect your interviewees being experts at research & development. It’s still pretty much your job to understand the customer’s life, environment, workflow, and uncovering pains and potential products.

The problem I’m finding with sales safari (and I’m a 30x500 alumni) is that you can still be building a product in a vacuum. You may not have enough data to decide to go creating a product after reading forums and blogs. So you can spend months creating an initial product only to find out the pain was not that strong after all. That can be alleviated if you take the time to talk to members from your audience.

So in my opinion you should combine both. Do some sales safari to get an idea of the pain/problems/existing products, then go deeper with members from your audience to understand them better. And at the same time validate that you can provide a solution.


#13

What? Are you sure we are talking about the same Amy and the same 30x500 course?


#14

Thanks for your feedback Reshal.

I remember feeling really excited when I read that post. I even tried to put that strategy in practice (only with signups, never added the “Pay” button).

I didn’t work for me. Getting some traffic to a landing page never felt reliable enough to me.

What if people that came there are not from my target market?
What if the number of people is too small to consider conversion rates reliable?
What if the design or copy of the page is the reason that makes people not convert?
What if the solution offered requires the users to test it before they can decide if they want to purchase or not?

Some months later, I found this great post by Dan Norris. It specially resonated with my experience:
http://wpcurve.com/is-startup-validation-bullshit/


#15

Yes, we’re talking about the same course.

The idea of the creative atom is to get something out fast and get feedback on that, but you still need some time to build a product during which you run the risk of being isolated from feedback.

What I’m saying is it can’t hurt to be proactively reaching out to your target audience to get more feedback (hence the customer development interviews that Amy doesn’t like). People reacting to what you’re building (be it blog posts, ebooks, software) can help but often times it’s not enough to get enough feedback, so you should reach out rather than avoiding people.


#16

Thanks for your insights @ShawnArnwine.

My actual process of finding product ideas is:

  1. Defining some requirements I want my product to follow (SaaS business model, minimal interface, extremely easy installation/integration, revenue from day 1, that can be built and sustained by a one or two people team…).
  2. Finding a clear target market that is already paying for similar products (Shopify store owners…).
  3. Finding an existing task that my target market is dealing with (customer support).
  4. Thinking about a simple way to improve that task (offering easier access to relevant data about the customer).
  5. Thinking about a really simple solution that fits the previous criteria and that I feel excited to build and proud to market (v1 of my possible solution).
  6. Research the market to see if there’s people trying to solve the same problem, and if I still think my product can offer something different/better.

Step number 6 is the one paralyzing me now.

My gut is telling me that I should build a minimum but polished version 1 and start the iteration process based on the reactions of customers respect the product.

My head is telling me that I should get some feedback from potential customers first, but I don’t find a strategy that I feel will give me reliable enough data.


#17

Good point @tommy_jarnac.

I’ve been tempted to join 30x500 a couple of times. Amy’s approach makes a lot of sense, at least in theory. I guess it makes my rational mind happy to think that I can remove so much risk only by observing methodically.

My problem with Amy’s approach is that I’m only interested in building a product if it fits certain requirements. That can be incompatible with starting with a problem and having to completely adapt yourself to solve that problem.

I like how you describe a possible balance. We’re often too worried about having clearly defined strategies we can follow step by step, but maybe it’s ok to organically combine observational market research, customer interviews and simple experiments.


#18

How have you been trying to get traffic then? Because it’s exactly important to get the traffic from your target market, and Twitter has been a really good companion at achieving just that.


#19

There are obviously a few ways to slice this and different approaches that will work…and I agree that there isn’t a hard and fast way to figure this out. Just to further clarify what has currently worked for me, albeit a small win thus far, has been a combination of the finding a pain and talking to people to further refine my solution and determine if they will pay for it.

Instead of coming from a solution first approach (which always tripped me up), I decided to find the pain and determine if it was worth solving. So I went out to applications that I currently use and scoured their forums looking for a consistent pain point.

A great place to start is in the “feature requests” category, most larger applications have this.

While checking through the Box.com forums I found a request that had been out there for years and that Box doesn’t seem to have on their roadmap. So I marked this down as a possibility. About a week later I happened to have lunch with a friend who works for Box and we were talking about ideas…and he said that he knows of Box customers have the exact problem I found…he had no idea about the forum request. That’s when I knew I had to look into this further.

So I set up a landing page to capture emails and posted the link in the forums. I got a total of 50 sign ups in a matter of days. Not a ton by any stretch, but enough to see there was interest. I then emailed each one individually and had conversations with them…mostly over email (which isn’t ideal) but then a few over the phone.

One other thing to keep in mind, if you do explore platforms, make sure people are paying something to use it or that you can get to their paying customers. Box for example gives so much away for free that the challenge for me has been identifying paying customers, but I’ve slowly managed to do it.


#20

“If you don’t do your job getting real data, you might as well get data that has a high likelihood of being false to fill out your spreadsheet.”

Yo, everybody. Hi hi hi hi.

The 30x500 process is not as Tommy is describing at all.

The 30x500 process is exactly the opposite of working in a vacuum: You study the audience. You interact with the audience by helping them in a natural setting (e.g. a forum, like I’m doing right now). You teach the audience. You get them to sign up for your mailing list. You find & foster natural sources of data — when the audience is doing its thing — you don’t give them an incentive to behave differently, and you don’t rely on faulty self-reporting. (There’s a reason social science research studies rarely rely on self-reporting. It’s because it’s very low accuracy.)

If you ask people “What’s your biggest problem?” or “What’s the secret to your success?” you will almost invariably get different answers than when you observe.

For example: Tommy says his problem is 30x500 taught him to work in a vacuum.

Meanwhile, observation says: Tommy, you took 30x500 >2 years ago and didn’t reply to almost any of the “do this and share your results” homework threads. Of 12 weeks of class, I only see your work for exercises 2, 4, 5 and 7.

No denying that 30x500 has gotten much much better since then — and, as an alumnus, you have access to all the new lessons for free — but even in 2012, the process still focused on starting with real data (Safari) and ebombing, rinse and repeat.

If you feel like you’re in a vacuum, I’d advise you to go watch and try the new 30x500 Bootcamp video lessons that show clearly the entire process and demonstrate the specific Safari & ebomb techniques (again). You may find that after all this time, you’ve misremembered.