Email Print


Nick Kozak for the Toronto Star
Case Studies
Posted: March 12, 2013
by Jacob Serebrin

Toronto’s Canopy Labs brings big data analytics to small, medium business — needs scale

Vital Stats
Canopy Labs
Wojciech Gryc and Jorge Escobedo
Customer analytics
Years active:
Data mining analytics platform for medium sized businesses
One of the company’s founders is a former Rhodes scholar; the other has a Ph.D in Theoretical Physics and conducted research on string theory.

Bringing big data to small business is a huge job.


Toronto-based Canopy Labs is only a year old, but they’ve already run pilot projects on their data-mining analytics platform with a diverse group of big names, including the Toronto Argonauts, the Canadian Opera Company and deal-of-the-day website WagJag.


The early success has come fairly easily, but further expansion means increasing overhead, and if the company’s strategy for building out it’s platform isn’t directly on the mark, scalability will be handicapped by the cost of development and winning new customers.


The company’s software takes a business’ customer data — everything from emails and sales records to call centre recordings — and analyses it using custom-built algorithms to reveal how loyal and satisfied individuals are feeling. It also predicts how soon customers will return, and what they’re likely to buy next, giving businesses the ability to laser focus their marketing strategies.


Canopy Labs CEO Wojciech Gryc compares his platform to the personalized recommendations offered on websites like Amazon and Netflix. It’s a telling reference; due to the high cost and specialized skills required, this kind of large-scale data analysis has previouslyonly been available to multinational businesses.


In order to change this, Canopy Labs has built an affordable, plug-and-play solution that any customer-facing business can use by simply uploading their data to the company’s website.


Their target market includes any business with between 5,000 and 100,000 customers — they’re charged $250 to use the platform.


Gryc says his “baseline for success” is if business owners who don’t have a background in math are able to use the platform easily and see results.


Currently, the platform allows users to upload data from Excel and the popular Shopify ecommerce and MySQL database systems. But the program much less accessible for businesses using other software, and there are thousands of other programs and different versions out there.


This is a problem, but Gryc believes he has a solution.


“We envision our platform having ‘apps’ where you can obtain support for whichever business tools you use by installing our ‘app’ and then having our software automatically sync with your data,” he says.


This is easier said than done. Because the platform uses so many different types of information, it’s not simply a matter of making sure syncing apps work with accounting programs. They also have to sync with business intelligence, email and database software. Each program uses a different format to store data, and while some are built to communicate with other software, many more are not.


This means that the time needed to build an application can vary dramatically, ranging from two or three hours to three weeks.


Gryc says the effort is worthwhile, because it gets results.


“Every time we’ve seen a business move to a more personalized approach, (there’s been) a growth in revenue,” he says, adding that some clients have seen revenue grow by as much as 200 per cent after adopting Canopy Labs’ platform.


These successes haven’t gone unnoticed. In early December, the company closed a $1.5 million venture capital deal with the Business Development Bank of Canada, PayPal co-founder Peter Thiel’s Valar Ventures, and with several angel investors.


They were able to attract the investment largely because the market for business analytics software is growing rapidly, reaching an estimated value of $1.875 billion US in 2011 — an increase of 10 per cent from the year before — according to business intelligence firm IDC.


But that doesn’t help the scalability issue, and the problem goes beyond building apps for the multitude of available business software — the company can’t simply code a program once and walk away. Updates to business programs can change the way they store data, meaning Canopy Labs’ apps will also need to be updated as new versions of the various accounting, database and email software suites it supports roll out.


In other words, finding more customers means supporting more programs, which requires additional staff for the increased coding load. Overhead will increase with sales growth — a death sentence for companies attempting to scale up.


So, to combat this problem, Canopy Labs is conducting market research to find out which business programs are most commonly used. One thing they’re already finding out is that despite the multitude of options, most businesses are using the same programs. This means a little work could go a long way when it comes to making the program accessible to as many

clients as possible. They’re also tapping their VC funding to hire additional staff to meet increased demand.


“There’s a growing recognition, whether it’s by companies or organizations or even, to an extent, individuals, that they can add little bit of scientific rigor to how they make decisions,” says Jacomo Cormo, Canada research chair in information and performance management at the University of Ottawa.


He adds that the biggest roadblock for businesses wanting to put data mining to work is finding expertise, because the majority of people with the skills to do this kind of modeling are already working at large companies like Google and LinkedIn.


That gives Canopy Labs and its open-box software a competitive advantage. Gryc says there’s at least 90,000 businesses in their target market worldwide — medium-sized enterprises with a few thousand customers and low price points. For these companies, being able to personalize messaging based on data is particularly beneficial.


Now Gryc’s challenge is reaching out to them, and making sure his software is compatible with their systems when he does.


“We’ve picked our problem to solve,” said Gryc. “We’ve tested it, we’ve validated it, so we’re going to run with it.”




As Interviewed by: Tom Henheffer

Big data is mostly used by large corporations, government agencies and NGOs — Canopy Labs is trying to bring it to the medium-sized business market. But they need more foundational work to really achieve success. The key is identifying their target market. If they think it consists of 90,000 businesses they’re defining it far too broadly, and that’s why they need to keep customizing the program differently for every company that comes along. So, Canopy Labs should shift gears, maybe hire a marketing expert, and do more research on their potential clients. They need to pick a sub segment — perhaps mid-sized banks in the U.S. or well established, medium-sized Canadian online retailers — and build one program that works well for the majority of companies in that sub segment. But working well doesn’t mean generating reams of data, they need to remember clients won’t be as tech-savvy as the company founders. The key will be creating a simple interface that provides actionable data clients can use to increase profits. Once they have this, they can dominate one niche and expand outward from there.

by Don Ludlow

This is a fascinating company with a great offering. Data analytics has been the domain of major enterprises like Exxon or Google — bringing an affordable option to small and medium-sized businesses creates a huge opportunity. But it’s important companies like Canopy Labs don’t develop a technology and just race off to sell. You need a base program that can be built out and scaled. Volume will be key for success, and they can’t achieve volume if they’re constantly customizing for every new client. This means doing research and beta testing to isolate common client needs and discover which ones fit into an open-box solution. Canopy Labs is already testing, but they should ensure they do so with the mindset that analytics algorithms are just a kernel of what they’re selling. The user experience makes the whole package, and they need to make that experience deliver, in an easy-to-use way, the information clients need to better understand customers and increase profits. Otherwise, it’s no problem for someone to come along with a similar — maybe even inferior — product with a simpler, more effective interface and corner the market.

by Gary Isaacs - Cisco

Canopy labs have the chance to become a major company, but they’ve fallen into a very common trap. They see the huge number of small businesses in Canada and the U.S., but forgot just how difficult it is to reach them. These companies are incredibly diverse, and they won’t all want Canopy’s software. In other words, Canopy Labs needs to do more market research. They have data sets already, and that’s a good start, but they need more than feedback from beta testers — they need to know who their target market actually is. It might be small hotel chains or travel companies that don’t yet have the in-house analytics of bigger players — it might be something else entirely — but it certainly isn’t every small business with more than 2000 customers. Canopy Labs needs to realize this and focus in so they can better tailor the software to customer needs. It’s premature to hire a VP of sales until they figure this out. They’d be much better off with a VP of marketing so they can understand exactly who it is that will want their software.

by Knud Jensen - TRSM