One of the anathemas of the Lean Startup is the Perfect Product(TM) :). I’m sure we would all like to buy the perfect product but should a company build it? Well generally no.
“That’s great but can I have it in blue with Alcantara coating?”
If we digress into market segmentation, a perfect product will be one where the target segment is very, err… targeted. In marketing, you might call this a ‘segment of one’. We are all unique so a product that is truly perfect will be perfect for me and not perfect for everyone else (hence the market segment has one person in it). So, what is the issue with this definition of a perfect product? The primary issue is there are 6+ billion people out there and each segment of one has a different perfect product and this would be prohibitively expensive for any company to build products for. Of course, if it is a product that the needs of the user are pretty common e.g. toothbrushes, it is a bit easier to generate a product that is ‘almost perfect’ for everyone but even there, people need different sizes of brushes, the hardness of bristles, colors etc. So coming back from the digression into market segmentation, a couple of the arguments against building the perfect product are:
- It is very expensive to build a perfect product as you need to understand the needs of the user intimately.
- It is costly to market the perfect product as you need to find your potentially very small niche. Finding this niche and marketing effectively to it is expensive and probably only makes sense if the absolute profit margin is very good (e.g. selling houses for commision).
- … and, this is what I really want to talk about, you may never actually get the product out as it is never done (never quite perfect).
This striving for perfection is bad as you may fail to release the product and when/if you do release your product, any error in your understanding of the requirements of perfection will jeopardize the likely uptake of the product. Further, the competition may have mosied in and pitched their product on your target market segment. So back across in Lean Startup land what do we release? Well, the key term is Minimum Viable Product (MVP). What are some key characteristics of the MVP?
“It’s in the name…”
- [It is a PRODUCT] It is an end-to-end product (i.e. it is not a survey or a powerpoint deck). It has to be something a real user/customer can use in the normal use case for the product. Example: if the product was a photo sharing site, the user must be able to log in, select photos (which must come from somewhere) and share them with whomever they intend. It would be no use if they could only log in a view the photos but not share them, that is not end-to-end for a photo sharing app.
- [It is VIABLE] If there is something that would stop the user using the product then it is not viable. Running with the photo sharing example, if it was a photo sharing application for real-time photo sharing, then one might expect it to be available on a device that you have with you all the time – e.g. a smartphone. If you only had a desktop application that ran on a PC, this would not be viable.
- [It is MINIMAL] This is where there is often a lot of debate. To use a frivolous example, if the product is to brush teeth, the toothbrush fits the bill. You do not need to design a toothbrush that has a nail clipper attachment. To go back to the photo sharing example, if it was about real-time photo sharing, it is highly likely that you want to be able to take photos on the smartphone and share these. You need to integrate the camera into your app right? Wrong, minimal would be to let the user use the camera app built into their phone to take the photos and have your photo sharing app access the picture library on the phone to share those photos. Is it pretty, no, is it convenient, no, is it minimal, yes. I would add though, if the minimality of the app makes the user not use it, then it is too minimal and not viable. So if the activation energy of taking a photo with the camera app then switching to the sharing app is too great (say 80% of users don’t use the sharing app as a consequence) then you need to revise the view of minimal.
To dwell on MINIMAL a bit more, if you don’t know that 80% of the users don’t want to use the sharing app MVP because they need to switch between the camera app and the sharing app, TEST it on real users. Find a few willing guinea pigs, ideally real potential customers and let them use the app and get feedback on what works, what doesn’t work for THEM. Remember you are building the app for your customers, who may or may not also include you. Don’t fall into the trap of refining the product until it is perfect for you… there’s that naughty word ‘perfect’ again :). David Aycan wrote a great post at Harvard Business Review on ‘cutting the fat not the essence’ of a product.
“I estimate it will take 10 years and 1 beeelion dollars to build!”
What if a feature is very expensive to develop either in time, money or both? The Lean Startup would suggest you ‘concierge’ the feature e.g. with photo sharing, perhaps you have some cool feature that analyses the photos you are taking and gives you recommendations as to which forums you should read on the web. This could be a hard computer vision problem and take months to develop. Should you just build it anyway, nope, concierge it first to see if the users find it compelling. What does this mean? Get a human (you) to perform the recommendation task manually and slipstream that process into the app. Test it out on users, evaluate if it is a ‘valuable enough’ feature for them and the company before plowing seed funds and time into developing it for real. Perhaps there is a different product you discover you should build in this process of getting real user feedback and you may want to pivot to that. It is cheaper/easier to make (big) changes to the product at this early stage than after the fully featured product has been completely built and released.
“I didn’t really want an aircraft carrier, I just wanted a yacht…”
One other caveat from the MINIMAL argument, perhaps the unique selling point of your product is the recommendation of forums based on the photos you take. The product is not minimal if you do not have this feature in there (whether that is in concierge mode or for real). MINIMAL is not an excuse for releasing a product that does not provide your unique take on the value proposition. Why would you release a photo-sharing app to compete with Instagram for instance? Not all products are a completely new category, so it is fine to release a photo-sharing app so long as for your target market segment of customers, there is a value proposition that is going to differentiate you e.g. it is cheaper than the alternative, has a unique feature that makes it more suitable for that niche etc. Perhaps making Instagram for microbiologists requires some features that are not in Instagram and that market segment of microbiologists who share mobile photos is a) big enough b) reachable c) profitable enough for you to build MicroBioInstagram :).
“Please eat my lunch why don’t you?”
There are some critiques of the Lean/MVP approach: sometimes, revealing your product to a market in an MVP stage may give the key business idea away and others will copy it and out-execute you in terms of delivering it to market. You should ask yourself whether your product/market has sufficient barrier to entry from competition if it is this easy to copy. Consider patents, copyright and other forms of intellectual property protection in this sort of market/product. This is not a panacea as not everyone respects IPR laws and it is expensive to get and defend patents etc.
“At last we will reveal ourselves to the Jedi”
(From Star Wars The Phantom Menace‘s Darth Maul)
If you are worried about revealing your product the world early, think about running small pilot studies and private betas. You can get your users to sign NDAs if you really want. At some point, you will have to come out of ‘Stealth Mode‘. I’m a firm believer that ideas are cheap it is actually executing and delivering products that make the difference but that is a completely separate post.
“Won’t it damage my brand?”
Some people claim that minimal is also half-baked and would cause damage to your brand. Would Gucci try out cheap digital watches as an MVP under the Gucci brand? No, but this is not an excuse to not try out the MVP approach – think about using an external company/brand to test out the idea/MVP – gauge the market interest before really doing damage to your brand if you are concerned.
What women want. What customers want
So to recap on the key message. The people who really know whether your product is any good or not are your users/customers. Show them your MVP as early as you can, don’t tinker with it or build unvalidated features. Don’t mistake this for asking customers what they want. Henry Ford famously suggested that if he asked what his potential customers wanted, they would have asked for a faster horse, not a motor car. Build the MVP of the motor car and get potential customers to use it and build on that feedback to build the next version of the product. The world would have been very different had Ford sat around tinkering his car until it was perfect and never got around to releasing it – perhaps the ‘Faster Horse Co.’ would be the dominant provider of transport.
Leave a Reply