Prototyping · Minimum Viable Product

Time to build that first prototype, your advice and suggestions?

Nimesh Thakrar

June 29th, 2013

Hi folks,
I have reached that stage where I want to get a prototype built, but want to make sure am taking the right approach. It's for an e-commerce site and my USP is that products are 3D printed. Looking for advice with regards to do's and don'ts, and ideally a approach to follow for UI, design and tech. I am aware of some of the 'theory' (e.g. from the Lean Startup), but eager to hear from others' real world experiences. 
Bear in mind that I'm working solo so will be hiring freelancers for now.

Eric Rogness Technical Product Manager

June 29th, 2013

While Big Commerce and Shopify are probably not flexible enough for your needs, do not try to build an e-commerce platform from the ground up: there are platforms of varying capability, from Prestashop and WordPress, all the way up to Magento. Your limited resources should be invested all in front end stuff: properly thinking out the UX, wireframing, creating slice-ready designs, and implementing them in a HTML+CSS theme, to make it your own. The special stuff you want to do can be modularized, either stuff you can plugin off the shelf, or append your own work.

When it gets to the actual UX stuff, I'm sure you'll have lots off questions worthy of discussion, but I'll leave that over till there's more detail and another thread.


June 30th, 2013

In terms of implementation, I would completely separate the WebGL functionality from the e-commerce functionality. The customization in 3D would just output JSON ideally. Then the e-commerce site just needs to be able to 1) embed the WebGL editor and 2) capture the JSON to update the product page / cart accordingly.

I know shopify lets you customize your site's theme and run arbitrary javascript on the page. So it's definitely doable. 

So start with the WebGL editor and work backwards. E-commerce, as people have mentioned, is a solved problem. Start with the part that actually requires custom development. And build it without any hard dependencies on any particular e-commerce solution.

Robert Clegg

June 29th, 2013

You must define/know your customer completely first. Create an imaginary profile of who this person is detailing the psychological drivers that compel them to do what they do and hence use your sight. This uncovers the value they see in your product and forms how you design your sight to meet their deepest needs. 

Not doing this often shows up in the classic problem of "features vs. benefits". Meaning us product guys love to talk about the features of our products instead of the benefits it can provide for the customer.

Duane Nickull Chief Marketing Officer, Co-Founder at Cheddar Labs

June 29th, 2013

UX testing 101 - ask people who did not work on the product to test it.   We use the MacKenzie-Nickull Architectural patterns Metamodel to capture user epics (basically a plain language description if how a user interacts with your software).  The metamodel then allows the pattern to be turned into a solid set of engineering requirements.

Shahab Layeghi Software Professional

June 30th, 2013

What do you mean by "a custom WebGL build?  Also note that WebGL is supported in most new desktop browsers, but not in all mobile devices yet.  So if mobile is important for you and your time to market is short, you might have to rethink WebGL part.

Rick Nguyen Cofounder @ Spot Trender

July 7th, 2013

Hey Nimesh,
This is how I would do it. The benefit of this approach is speed and extremely low cost.

-use wordpress and google "wordpress ecommerce theme" to build the prototype. Buy the theme and install. It's simple enough I think you can get it done for less than $200.

-I googled "wordpress webgl plugin" and found some pretty awesome guide/plugin, such as this:

Hope this helps.

Nimesh Thakrar

June 29th, 2013

Thanks Craig - I should add that a significant part of the site will be product customisation in 3D. So a WebGL embedded 3D editor or similar. I want the editor to do specific things that will require a custom WebGL build. So reckon I should build that and make sure it integrates well with big commerce, as a first step? 

Craig Green

June 29th, 2013

Are you building a website or a product? If a website, just get a site from a place like big commerce, etc. they've solved all the hard problems already. Craig Green 703.864.3502(c)

Eli Chan

June 30th, 2013

I would recommend embracing the Lean Startup methodology. Your first "minimum viable product" need not require any web site at all. Perhaps you could find some early adopters and simulate your offering via Legos?

Brahm Singh Eng Head - Deep Search at Quixey

July 1st, 2013

I can share my personal experience. We started out to build a virtual teacher - a learning system which will teach you topics which you are weak at. Pretty daunting task.

We started off without a site - using google docs to give out tests, record responses, simulate what a real teacher would do.

Got some learnings - people don't want what we were building.
Ditched the idea, tried a variant, got a decent response, iterated, got better.

What we got from all this - don't build unless you know that your user really loves what you are starting out to build. Use mocks - say Legos like Eli suggested. Don't be afraid to test out such simple mocks.