Week 3 Summary: Designing the Smart Home Matrix Proof of Concept (PoC)

This week, I designed what the initial proof of concept will look like for my smart home matrix.
7 March 2022

It’s time to start designing, life has been hectic lately so this is a bit delayed. I have a good idea in my head about what I’ll be building, now to break it down into bite size pieces.

Defining the features of my Smart Home Matrix

In the Week 2 Summary I covered what features I’ll include in my proof-of-concept build. 

While being pretty straight forward the end-goal of this PoC is to have two things;

  1. A working prototype, that would provide immediate benefits to users (accurate information collated into a single source)
  2. Allow me to start gathering feedback on what I could improve / include going forward (real tangible feedback)

PoC Features

And here’s the features that I’ll start with;

Users will be able to;

  1. Select a product and find products that work with it natively / directly (i.e. you don’t need another device or piece of software to join the two products)
  2. Find a product based on certain criteria (e.g. Find all smart locks that connect via BlueTooth)

I’m going to add that this PoC won’t include all smart home products initially, and will instead focus on popular products first to give an idea on how this matrix works.

The data harvesting required to enable these use cases isn’t too difficult, as manufacturers tend to list key information about their products' capabilities online.

Challenges with this PoC

Finding the right tool proved harder than expected. I struggled to find a no code solution that will allow me to create this matrix, I checked out several tools.

  • Stackerhq (Spreadsheet into App | Expensive)
  • GlideApp (Spreadsheet | Costly)
  • Jotform (Only 100 submissions Free)
  • Softr.io (Limited Airtable Rows | Costly)
  • Bubble.io (TBC)
  • AppGyver (Too mobile focused)
  • Coda.io (Not for external use)
  • https://www.involve.me/pricing/ (100 free then like $25 per 1000 submissions)

While I initially thought this was a pretty straight-forward PoC, finding the right tools has proven to be quite time intensive. Also harvesting the data required as there’s thousands of products on the market right now, is going to take a while.

To combat this, I’m going to start by focusing on major product areas (smart speakers, smart locks, smart lights, etc) and brands (Google, Amazon, Nest, Samsung, Philips, LiFX, etc) only for now.

As for which tool I picked, I’ll go over that later on - but it was actually bubble.io.

What the Smart Home Matrix will actually do

Now I know the features I want to focus on implementing, I had a bit of a brainstorm on how the matrix should work. Above all else, I wanted to create the best user experience (UX) possible.

As the PoC runs on a large amount of data, it’s all about displaying it in a usable manner. I won’t jump down a rabbit hole of turning data into information into knowledge, you can read a short Cambridge paper on the idea here that explains it at a high level. 

Turning data into information

Now the question becomes, how do I turn the data I collect into information that’s useful to anyone using the matrix. In particular, how do I display the data in a manner that actually provides value to the user.

Here’s the top two ideas that I came up with;

  • Option 1: Find what a user would like to know by asking a series of questions (one or more) to find a relevant answer (using simple / advanced logic)
    1. E.g. “I’m looking for products that work together” or “I’m looking for products that connect via X (WiFi, Bluetooth, Zigbee, etc)”
  • Option 2: Have all product data in filterable spreadsheet, allowing users to filter certain fields (simple but not very user friendly)
    1. E.g. Imagine a giant excel spreadsheet.

User Experience (UX)

Option 1 is what I’m leaning towards, I’ve used similar tools in the past and it’s generally been a good UX. 

I can see a user picking “Find me a product based on X” option, then through a series of questions narrowing down what they’re finding. Below I explain how I see this interaction happening.

Option 2 on the other hand, might be useful to some but quite intimidating to users that aren’t familiar with how to properly sort and filter data.

Mock Up

Based on option 1, I created a quick mock-up on how I see this matrix looking (using Miro), which shows the basic user journey someone will go through.

Of course the user experience will change slightly depending on the tool(s) I use, as I won’t be building / developing this matrix with much if any custom code. I’ll be limited to the tool I pick.

Similar apps on the market

Now I know this product selection / product recommendation has been done before - mainly in the eCommerce industry, so I decided to do some research into the competition, both tools available and sites using them.

Tools on the market

To no surprise, I found several companies offering similar matrix style software to what I’m trying to build, but not exactly what I had in mind. I used several terms to find possible tools around “Product Recommendation”, “Quiz”, “Engine” etc.

Here’s a few tools I found.

  • Revenue Hunt | Product Recommendation Quiz | eCommerce Focused
    • If we look on “builtwith.com” we can see (roughly) 7,419 websites are using this product. Through further investigation, it seems some of these sites no longer use it.
  • Expert Village Media Technologies | Personalized Recommendations | eCommerce Focused
  • And more… 

I won’t bore you with this research as it was pretty straight forward.

  1. Identify key market tools
  2. See what they can do (are they fit for my purpose)
  3. Are they actually being used, find live examples

A problem with these tools is they’re quite restricting, as in, it’ll be hard to customise them to my use cases and beyond. I’d essentially be stuck with the way they’re programmed and can’t really alter the code.

This was an easy no-starter for me, while these apps are probably great for eCommerce stores (pretty obvious by their marketing), this wasn’t a path I wanted to explore further. I needed something more customisable.

Other ways to complete this matrix would be to use several tools together

I then had the idea of hooking up several tools to create my “matrix” such as

  • Typeform + Airtable + Zapier
    • This would allow me to store data (Airtable), use logic to guide the user (Typeform), and then tie both together to produce an answer (Zapier).

After more research and reflecting, I realise adding more tools will add unneeded complexity and increase development time, not to mention if something breaks - trying to figure out where it’s broken.

Creating a PoC matrix without breaking the budget

Another thing I found when researching tools is their cost. Saying they can be quite expensive for my use cases, would be an understatement. For example let’s look at the Typeform + Airtable + Zapier combo - that could potentially work (rough research not in depth validation)

  • Typeform is priced (Basic option) at £21 a month for 100 responses a month.
  • Zapier is priced (Free) with a limit of 100 tasks a month
  • Airtable is priced (Free) with limit of 1200 records per base

Given the above, if I just wanted to provide 100 readers a month with a one-time answer to their enquiry that’ll cost me £21 a month or (£252 a year).

Let’s say instead I want to give 1000 people an answer to their question, now we’re talking;

  • Typeform (£41 per month) for 1000 responses a month
  • Zapier (£28.53 per month) for 1500 tasks a month (no 1000 option available)
  • Airtable depending on size of database, might need to use larger database (£12 per month or Free)

That’s now £834.36 per year total. 

Honestly this just isn’t feasible for a fun project that isn’t looking to profit initially and is designed to help others - I don’t want to limit how many people can use my matrix.

Road block, first hurdle

I couldn’t find an app, it seems the matrix I have in mind and the recommendation or guide logic doesn’t seem to exist in a no code tool (in a simple way). Posting on reddit and makerpad didn’t result in an answer, instead people advised me to look into using tools such as Bubble.io and Adalo.com. 

I didn’t see a need for this matrix to have it’s own dedicated iPhone or Android store app (I know Adalo can do web apps, but wasn't convinced), so I decided to look more into Bubble and alternatives.

Before I jumped into bubble I wanted to see if there was any rivals, alternative tools that could be useful - which there didn’t seem to be any.

  • Retool - Only really for internal teams, doesn’t allow the public to use the built app.
  • Wavemaker - Way too expensive at a starting price of $500 a month
  • AppSheet - Potential but demo apps didn’t seem too promising, low price point but no free version to try.
  • SpreadsheetWeb - The $25 starting put me off but it’s a potential for later on
  • Betty Blocks - Have to sign up for the price

If I’m missing a tool that competes with bubble please let me know, I can’t really find direct competitors. Probably one of the reasons why Bubble.io raised over $100M in their Series A funding.

Initial Bubble.io Impressions

What’s better than free when trying to figure out if something will work.

After my initial signing up, I completed the first few tutorials recommended. While Bubble might be overkill for what I’m trying to do - I honestly couldn’t find a no code or low code tool alternative that could bring this matrix to life. 

Not to mention from all I've read, bubble might become one of the main no code tools I use for future projects.

Before diving into creating my matrix, I carried out a few more tutorials that I thought would help me in some way.

Aside front he guided tutorials, I also did parts of the demo projects like “Build a ticketmaster clone”, which I found helpful in certain areas. I think I know enough to at least attempt the PoC for the matrix for now.

Now to figure out the larger part of this project, collecting and manipulating data.

What data I need to capture and where I’ll be storing the data

Next on my list of things to do is figuring out what data I need to capture. Though it’s probably best I learn a bit more about database design before collecting anything.

Googling, I stumbled onto several helpful sites with one I found particularly helpful, Vertabelo, which has some great articles on all things database related.

I found an article that was semi-close to what data I’d be capturing which was “Data model for online musical equipment shop”. It was helpful to determine data structure and I’ll start modelling it next week.

Collecting data is going to be the least entertaining part of this project, but without it my matrix doesn’t exist.

Building a quick PoC

Now I have the tool selected (Bubble.io) it’s time to start building. Next week I’ll hopefully have the following created.

  1. Database design (structure)
  2. Sample data (for testing purposes)
  3. Working prototype (using the sample data)

Getting a working PoC going will be a great achievement, it’ll validate that my idea can be brought to life, now the easy part to see if my research is right - that this is a product that will help people find the perfect smart home products.

Subscribe
Notify of
guest

This site uses Akismet to reduce spam. Learn how your comment data is processed.

0 Comments
Inline Feedbacks
View all comments

Latest Content

29 March 2022
Week 4 Summary: Data to collect

This week it was time to understand what data I needed to collect, which turns out to be a lot more than I expected.

× Read More ×
7 March 2022
Week 3 Summary: Designing the Smart Home Matrix Proof of Concept (PoC)

This week, I designed what the initial proof of concept will look like for my smart home matrix.

× Read More ×
28 February 2022
Week 2 Summary: Planning the Smart Home Matrix

Completed the initial planning phase of this project, from figuring out what I'll be build as Proof-of-concept to the user journey and more.

× Read More ×
23 February 2022
Week 1 Summary: Brainstorming and validating ideas

Finding a problem to solve starts with an idea. Trying to find a good one however isn't easy, which is where brainstorming comes in.

× Read More ×
10 February 2022
Launch useful iPhone apps faster, by simply double or triple tapping

Launch apps on your iPhone like your camera by double or triple tapping the back of your phone.

× Read More ×
5 February 2022
Use your iPhone with one hand, with the reachability feature

When you only have one hand available and need to reach the top of your phone. You should use the reachability feature built into the iPhone.

× Read More ×
26 January 2022
Using the inbuilt trackpad on your iPhone and iPad

Use your iPhones inbuilt trackpad to move the cursor to make it easier to fix spelling mistakes.

× Read More ×
24 January 2022
Use live text to copy physical text straight to your phone

Use live text on your phone to quickly copy physical text directly to your phone. It's easy to use and really powerful.

× Read More ×
17 January 2022
Text Replacement - Type faster and reduce spelling mistakes

Learn how to quickly type your email, mobile number or even home address online. Saving you time and reducing the chance you'll misspell it!

× Read More ×
linkedin facebook pinterest youtube rss twitter instagram facebook-blank rss-blank linkedin-blank pinterest youtube twitter instagram