San Fran says go
Silicon Valley drops $34m on Kiwi startup PredictHQ
PredictHQ’s founders did not predict their giant Series B round would happen, but they’re not complaining. The Auckland-based software startup has just closed a US$22 million ($34m) raise, led by one of Silicon Valley’s marquee venture capital outfits, Sutter Hill Ventures, in its first major investment downunder. Ownership stakes and a post-money valuation have yet to be disclosed.
Co-founder and chief executive Campbell Brown says Predict HQ wasn’t looking for a new investor. After it raised US$10m in a Series A round nearly 12 months ago, it was quite comfortable.
But then the unsolicited Sutter came calling — and the VC company is always unsolicited.
“They don’t have any contact form on their website,” Brown says.
“You only hear from them if they want to seek you out.”
Brown says a startup always has to be open to new money. You should always take an opportunity while the going is good. And given the chilling effect on the business environment from the coronavirus outbreak, he was glad the deal was nailed down over Christmas, before the crisis erupted.
PredictHQ’s software engine tells companies what events are going to happen — “demand intelligence” that helps them factor in the impact on their forecasting, pricing and planning.
When the Herald visits, Brown is scheming to watch the Black Caps play India in the third one-dayer. “I’m trying to soak up as much cricket as I can before I go back to the US,” he says, before asking where the third ODI is being played.
“I should know,” he adds. Or at least his software should.
PredictHQ tells clients like Uber, Qantas and Domino’s what events are happening and where, and how factors like the weather could impact.
To a lay person, the opportunity wasn’t obvious. Surely Qantas and its ilk know when, say, a Phil Collins concert and a big tennis tournament will happen in the same city on the same day, and how they will impact demand.
But, like many of the best startups, it was born out of its founders going looking for a service they needed, not finding one, then building it themselves.
In 2015, Brown and Robert Kern (now PredictHQ’s chief technology officer) were part of the crew running Online Republic, a site that offered online car and camper van rentals and cruise ship bookings.
A key problem was predicting demand. Online Republic was constantly caught out by public events or, more particularly, when several public events happened at once, especially around holidays — those weekends when there’s a big test, a festival and a concert running into each other.
Brown was amazed to find that while there were lots of forecasting tools that mined historic data, there was no global player who aggregated all public events in one, easy to access service.
Online Republic was flicked off to Webjet in 2016 — for a tidy $85m — and PredictHQ was born.
The startup began telling Uber which events were likely to draw its cars, and Domino’s where the hungry were likely to gather. The travel industry was a big focus.
Now, PredictHQ has moved into new areas such as “labour optimisation”, which it’s automating for one of the world’s largest retailers (which Brown won’t name, but which he says has around 14,000 stores in the US and some 27,000 globally, somewhat narrowing down the field of candidates.
“They ingest our API [ application programming interface into their data lake,” Brown says.
They what now?
The chain’s store managers used to waste a lot of time trying to work out how events would weigh on customer demand and how many staff need to be rostered to each location. Now the process is taken care of by PredictHQ.
The company is now also working with a “larger card provider” so that it, in turn, can help local retailers with their forecasting, from events that drive demand to nuances such as an event’s impact on merchants who are a little further away.
“We help businesses all around the world better understand demand, and what anomalies are impacting demand,” Brown says. “Whether it’s an Uber trying to get drivers in the right place at the right time, ahead of time, or someone like Dominoes, who’
[Sutter Hill Ventures] don’t have any contact form on their website. You only hear from them if they want to seek you out.
Campbell Brown, PredictHQ
needs to forecast how much pizza dough they need in a store before a major [College Basketball] game.”
The Kiwi company has no direct competitors at present, and Sutter Hill wants to seize the day. PredictHQ increased revenue 125 per cent last year, but the new Silicon Valley investor wants to see even faster growth.
Part of that will involve putting more bodies on the ground. Predict HQ currently has 45 staff at its Britomart office and another 20 in San Francisco. “That will be growing to around 80 here in New Zealand and around 40 in the US with an office opening in Europe towards the end of the year, too.”
He adds, “We’ve also got a massive product launch coming up toward the beginning of the second quarter, which is a correlation and prediction engine. It’s an opportunity that we spotted probably about four or five months ago, when we’ve worked closely with some of our airline and retail and transport customers.”
And what is a “correlation and prediction engine” when it’s at home?
“Basically, what we’re doing is helping businesses to understand what events impact them specifically, and how demand fluctuations apply specifically to them. A Qantas is impacted by events differently to an Uber.”
Brown won’t talk financials, but he says PredictHQ is only dealing with clients who can afford to pay hundreds of thousands per year.
Like a number of local software success stories canvassed recently by the Herald, including CIN7 and Flowingly, PredictHQ has spurned the world of low-margin apps and freemium software for SMEs in favour of aiming at the top end of town.
Today, PredictHQ is tracking around 25 million events across some 30,000 cities, and sales meetings involve Fortune 500 companies and lashings of data science jargon.
Its founders retain their quirky roots, however. Kern wears jandals around the office, while Brown shows off R2D2 suits that the pair used to wear at trade shows as a cheap, lowtech attention-grabbing gimmick.
“They were the best leadgenerators ever,” Brown says.