Mike Frizell is the founder of Paws for Life, a dedicated, online pet food delivery service. The model has found success overseas, but Frizell has successfully imported it into Australia – the company is turning over more than $10 million a year.
Frizell spoke to SmartCompany about his company’s success, and why analytics matter so much in helping the company reach its full potential.
We started in August 2011 with 16 different products and four brands, in Sydney. We started with three people – three cofounders – two have now left.
My background is engineering followed by management consulting. I was an investment banker…not in online pet food.
The business model is very similar to one that exists in the US. There are very similar companies in the US. We’ve moved into cats and want to offer the full range of animals.
Our market is still small – less than 1% of sales in dog food are online in Australia. In the US, Germany, UK, 10-15% are online.
We’re going to keep ourselves very busy keeping up with that growth. We’ll double our range in the next 12 months, just to make sure our customers have everything they want.
We want to make the customer experience seamless. Food delivered to the door, there when you need it, easy to get us on the phone, customer-experience driven.
If you look at the pet food market, you can segment it into two subsections: 50-60% of the market is supermarket food. And then the other 50% or so is 15kg bags from pet stores. Most buyers are female – and don’t want to carry lots. So buying must be convenient.
What’s unique about us, from a tech point of view, is you can buy your food at our site just once. Pick your dog food, and say you’d like a delivery every 10 weeks, and every 10 weeks, it turns up on the door.
You wouldn’t believe how seasonal some of the items are. One month could see 40-50% of annual sales of an item. If you miss it, it’s gone. So we look at Google Trends data to stay abreast of that.
We’re a pretty analytically driven company. So we use analytics tools, and a few others, to monitor the performance of marketing, of changes to our website, and to test the mathematical significance test of results.
A good example is this: We ran a test regional marketing campaign in two regions that had very similar characteristics – Wollongong and Newcastle. We tracked changes in key metrics in that area, then analysed data significance, and determined which marketing channel was more effective.
I find Google analytics data, if set up effectively, is very, very effective. Our head of tech has a PhD in applied maths. It’s useful because we have a very automated system.
To get real, insightful, confident decisions out of it is not difficult. To get indicative results is trickier, as is isolating whether an impact is significant.
But it’s not hugely difficult. I can do it. But having someone with that incredibly high level of mathematics like our tech head is useful.
Supermarkets don’t sell premium, expensive brands. They just don’t have access to them. People who own the brands don’t sell to Coles and Woolworths, for whatever reason.
We got access to those brands, but it was a very long process. It took a lot of working with suppliers to show we were good business partners, people who were going to be around for the long term, who cared about their brand and products.
Our relationship is now strong, but that didn’t happen overnight. It took 18 months. We had to show we delivered on our promises and were acting in the right way. It was very important.