How Tectonic Used No-Code Machine Learning in App Creation

Use Cases

A no-code AI case study.

Illustration by Pablo Stanley
Words by Dan Perry

Building products and services has never been easier with no-code.

And with the rise of no-code tools like Webflow, Zapier and Glide, there are now more opportunities than ever to create new businesses. Unfortunately, the often-quoted stat of 90% of startups failing will likely remain true and according to CB Insights, the key reason for failing according is that this is no market need.

Understanding your market and potential customers are key to business success and the best way to do that is through data. We find the blend that works best is to use a mix of qualitative and quantitative data to make educated guesses about what to validate.

We're big fans of data—especially as a tool for guiding fast decision making.

As Tectonic's appointed Non-Technical, Technical Person, I've always wanted to experiment with machine learning, but like any startup, time and resources are limited and we have to choose where to invest our focus. Even though we have a team of developers, they often have more pressing things to do.

Then a few months ago I came across Obviously AI: A no-code machine learning tool which immediately piqued my interest.

So we decided to take it for a spin for a client project. Let's call this client Poppa.

The Problem:

London is awash with different exciting pop-up restaurants, shops and experiences, but it's very hard to navigate, especially if you're visiting from another country.

The Solution:

Create a simple to use app that shows the best pop up experiences in London for tourists.

Potential Target Users:

Primary Users

  • Tourists visiting London

Secondary Users

  • UK visitors to London 
  • People that want to experience different cultures
  • Other Londoners

Potential Target Customers:

Primary Customers

  • Pop up restaurants
  • Pop up venues

Secondary Customers

  • Pop up experiences

Unbiased Validation

We believe that validating critical assumptions about your business as quickly as possible is the key to ensuring success. We do this with qualitative target customer interviews and data analysis of the market and the business.

However, with Obviously, data analysis from external data sources is possible without bugging our developers. This helps us derive insights about Poppa's target customers and most importantly, predict their behavior.

What To Do Next?

Like any startup, a key problem that Poppa has is that resources are limited, and they're not sure where to invest their time and capital.

"What should we focus on first?"

"What type of experiences should we begin with?"

"Which tourists spend more money when they visit the UK?"

The Poppa team had some anecdotal facts but wanted more evidence before they continued.

Getting the Data

There are many sources to get free public datasets to help your business needs. Some great examples are Kaggle,, the Bureau of Economic Analysis etc.

We found a great dataset on the number of international visitors to London from here

Excellent! So we cleaned up the CSV and uploaded it into the Obviously app.

Now, Let’s Start Talking to Our Data

Step 1: Analyze the data.

Serving a customer type who is already exhibiting the behavior you want—in this case spending money abroad—is a good market to enter. A reason for this is that you’re not trying to change customer behavior too much.

So the first question we asked was the “Average spend for tourists since 2002.”

Based on the results, we found on average that German's spend the most when they visit London. Then it's French tourists, followed by Australia. Depending on Poppa's business strategy, they could choose a few different routes:

  • They could focus on serving tourists from the country with the highest spend i.e. Germany.
  • They could choose to serve multiple European countries i.e. Germany, France, and Denmark
  • Or they could focus on Australian, tourists who may be slightly easier to market towards due English being their first language.

Another question I asked was “Which tourists spent the most over 10 nights in London in 2019?”

Our assumption was that after a certain amount of time, people who are spending over 10 days in the UK would need something to do—and pop up experiences are great experiences to try!

Ordering the data by country, we see some of the usual suspects (in the friendliest way) pop up. Spain, France, Germany, Italy with an average of 1 to 3 nights in London. These insights really allow us to start understanding what we need to test to see if there is a good business opportunity. 

What was particularly intriguing was that over the last 20 years most of the tourist spending was done in Quarter 1. Might this be because of the rise of cheaper flights after Christmas? 

This insight would suggest that catering to the bad weather in London could be a potential promotion and user acquisition strategy? Maybe the Poppa app could highlight venues that are warm and cosy, or take into account any pop-ups that have winter sales?

Using Obviously, we can analyze historical data to give us many more insight-driven assumptions to test than just going on gut instinct alone.

Predicting Customer Behavior

Step 2: Predict behavior based on tourists’ spending. 

Again, our hypothesis is that the bigger the spending habit (or the higher the intention is to buy) the greater the likelihood is that you could convince a tourist to try new experiences.

Given these predictive assumptions, we wanted to understand what was the highest impact on spending.

This is where things really get interesting: The number of nights tourists spend in London doesn't affect tourist spending behavior positively or negatively. This means that the length of time doesn’t affect a tourists propensity or likelihood to spend.

Based on this, Poppa could test the positioning of Poppa as a brand to allow tourists to try a new experience in London every day. 

Looking further into Obviously’s predictions we noticed that the key drivers which are directly proportional to the spend were the amount of tourists visits—which according to the site, is the number of time a person has visited London. 

The insight from this prediction suggests that the more times a tourist visits London, but not necessarily stay overnight, the more likely they are to spend.

To put it another way, the more times they've been to London = the more they are likely to spend.

This knowledge gives Poppa a great place to begin final assumption forming. They now possess data which suggests they should not focus on first-time visitors to London and instead focus on repeat tourists.

What Does This Mean for Businesses?

Using Obviously AI, we were able to quickly get data that allowed us to understand the market and generate some evidence-based assumptions to validate. We find that the faster you learn, the faster you can make better decisions.

Within the team, we have also been able to create better hypotheses that were based on actual data. The proving and disproving of these are higher value and let us create tests that will have a greater impact on Poppa’s business.

Machine learning has been a bit of a buzzword. And unless you were an experienced data scientist, it was difficult to understand the true implications to your business. With Obviously AI we’re now able to generate insights that would have been impossible to get access to— and use that to help businesses be more successful.

About Tectonic

Tectonic validates business ideas by getting unbiased answers from their exact target customers. To see more of what Tectonic does, visit their site. logo

Become the Data Scientist your team always needed.

Get Started

Get Started Now

See how no code machine learning can transform your business and change how you make decisions.