First Look

Suto: How to find out what to buy

A unique service aims to answer your product purchasing questions

texting 6497720551 79c434a2a0 b
Credit: Jhaymesisviphotography / Flickr

I recently got Listserve message (see my recent post) in which the lucky sender, Karim El Rabiey, wrote:

Two hours before I got the email that I won Listserve, I was thinking about how I was going to get my side project in front of some eyeballs. This seems like the type of opportunity that I should not turn down from the universe. It's called Suto (asksuto[dot]com) and it helps you search for products you’re looking to buy - via text message. Check it out if you're looking to buy something!

I love discovering things like this so I immediately tried it out. At asksuto.com you enter your question - I asked for a recommendation for a quadcopter under $500 - and give your mobile number. Within minutes I received a reply:

+14157797886: Hey there - thanks for texting Suto. There are 131 questions ahead of you that I need to get to. People usually get a recommendation from me within 6-12 hours. In the meantime, feel free to send me more details or any preferences you may have. I may even have a question or two of my own. -Suto 11:13 AM

The next day I got my answer and it was pretty good:

+14157797886: Hey there - I looked into the best quadcopters for under $500 for you. A really good choice is the Blade 350QX3 RTF: http://amzn.to/1IUNtUL Pretty fun to fly, and you can get some high quality footage if you have your own GoPro/camera. You’ll get 15 minutes of flight time, and relative steering, and GPS/Altitude hold for those windy days. Plus, a return home function! Another good option is the Bepop Drone by Parrot: http://amzn.to/1Jav2bv It comes equipped with a full HD 14 megapixel camera stabilized on a 3-axis framework for smooth shots. Also, you can pilot it right from your phone or tablet. You can plan your flights ahead of time, and let your Parrot fly itself. And you can even pair it with an Oculus Rift - how cool is that? Although there are a bunch of cool features for the Bepop, it is known to have occasional issues with control glitches. Let me know what you think of these recommendations - or if you need anything else! Cheers, -Suto 4:32 PM

Intrigued by the Suto concept and curious about where Suto might be heading, I got in touch with Rabiey with some questions. He replied:

Before I get into responses, I just want define a term I’m going to use: “Asks” are questions people ask Suto.

Me: How many people are doing the research?

There’s currently a team of 4 people that are responsible for the research and responses to Asks. However, there is an external group of product category experts that we consult with as a part of our research. They are utilized when appropriate and when there is a match to their specific product expertise (someone who knows a whole lot about headphones, or drones for example). 

How scalable do you think the service is? 

In all honesty - in it’s current state as a side-project - it is not scalable. Our biggest focus is going to have to be operational optimizations. How can Suto give better responses quicker? However, we have already noticed several economies of scale due to a few factors:

- There are about 30% repeat Asks right now that we can re-use answers for. Essentially people asking the same question (e.g. “What are the best wireless headphones under $100 for the gym?”).

- We can heavily control what Asks we get in our system via our growth strategy. We look to expand a product category at a time, get really efficient at answering Asks in that category, and then start growing into a new product category. For example, we’re currently quicker at answering Asks about headphones than we are about baby-strollers. This is because we’ve focused on getting more headphone Asks than baby-stroller Asks.

- We’ll never want to get rid of the human touch in Suto’s research. However, we have plans to build our technology to understand natural language Asks and do as much of the heavy lifting as possible via automated methods. We imagine an interface where where the person responding sees the Ask with suggestions already surfaced. And since they understand the “human context”, they’d be able to pick the best option(s) from the surfaced suggestions.  A [ugly] version of this currently exists in the form of an Excel spreadsheet :)

With these factors in play, we have a high level of confidence in the scalability of Suto. It won’t be easy, but it’ll probably be the most fun part for us to work on.

What’s your plan for making money?

When available, Suto uses affiliate links for products that it recommends and makes a commission off the final sale. However - not all products Suto recommends are affiliate links since we’re honestly more focused on giving the best recommendation rather than one that will make us money.

We’re also exploring options on how other business/services can utilize Suto’s technology and operations for their own purpose. In effect, imagine large e-commerce websites wanting to incorporate a recommendation engine for their catalogue.

I’m intrigued by the concept of Suto but as Rabiey admitted, scalability is a key issue. Even so, that may not be insurmountable if the service is underpinned by a strategic knowledge framework and its areas of expertise can be focussed. There are plenty of other companies with similar established “ask anything” businesses (e.g. Quora, AllExperts, JustAnswer, etc.) but Suto’s focus on providing product purchasing advice by texting is, as far as I can determine, unique (if you know different, please let me know).

So, given the vast pool of expertise reading this article, what should Suto do to enable scalability and drive business?

Comment below or to feedback@gibbs.com or tweet to both @quistuipater and @asksuto.

Join the Network World communities on Facebook and LinkedIn to comment on topics that are top of mind.
Must read: Hidden Cause of Slow Internet and how to fix it
Notice to our Readers
We're now using social media to take your comments and feedback. Learn more about this here.