MrWeb 21 YEARS FEATURE: Giles Palmer Interview
Continuing our focus on mobile insights, Brandwatch founder & CEO Giles Palmer talks sifting through 1.5 trillion pages of data; pulling apart the Model-T Ford; and why our industry could grow fivefold in 5 years. Excerpts from a podcast interview by Jamin Brazil of Happy Market Research.
Brandwatch is a social media intelligence company, helping clients to analyse and utililse conversations from across the social web - and employs 500 people. Prior to starting it in 2007, Giles co-founded a software company called Runtime Collective.
The full podcast of Giles' interview will be available in the next month. Please email email@example.com if you want to be alerted when it's live
JB: tell us about how your parents influenced your career...
GP: I had kind of a traditional 'stay-at-home mum and working dad' family. My dad had a very high work ethic, came from a working class background, a smart guy that did OK at school and got into quantity surveying. He joined a firm when he was 27, retired from the same firm when he was 55 and a Senior Partner. That said, he didn't put too much pressure on my sister and me to work hard when we were kids, in fact I often asked him for advice and he was reluctant to give me advice: he said 'you've just gotta figure it out for yourself', kind of thing. I went to a private school and a good university where I read physics, came out pretty young with no idea about anything and went into accountancy just to get a general business grounding.
I wasn't very good at it - my heart wasn't in it - and then I joined a few other industries, didn't really get on with any of those. I was a geeky kid, into computer games etc.. and the Internet didn't really appear until I was in my late 20s, when me and some friends started a company which we called Runtime Collective, building web applications for people. I think if I picked one thing I got from parents it would be that work ethic - I was kind of a lazy kid but I can do things reasonably well when I try, especially the science stuff. Now I work really hard!
My mum was incredibly organised and ran the household, so together they set us up as kids really well, provided a safe and loving environment. I think because I come from a middle class, very safe environment I've always been attracted to risk more than other people - there was never any sense of jeopardy anywhere.
Now I have 4 children including one stepdaughter - aged 16 to 20. I think I've taken the work ethic thing too far with them, put more [academic] pressure on them than my parents put on me [laughs]. I'm very impatient, in particular with my son who is happy doing nothing but online gaming. When he wakes me up at 5am screaming down the headphones, I lose my cool! But habits you pick up from your parents are very strong. They'll probably do the same thing to their kids and so on down the generations - that's my own guilt coming through from when I was a kid - it's one of those tragic things that you repeat the mistakes of your parents!
However... we have a low drama household and we all muck in to a degree - hopefully that will give them a solid footing. Since the start of lockdown, I've been really impressed with the way they have stepped up regarding their granny. She is now on her own stuck at home, and I'm really proud of the fact they are now actually calling her twice a week - because they see her vulnerability, in her 80s and on her own.
In the Pipeline
JB: give us some context re Brandwatch - what's it doing today?
GP: Firstly, we are trying to bring together some of the innovations that we've been working on for last couple of years behind the scenes, around NLP, universal translation and some of the Open AI projects that have come to light. Using these to analyse the enormous volume of social data: we've received something like 700m posts a day for 12 years - maybe 1.5 trillion pages. In that archive there are interesting insights, it's just How the hell do you sift the signal from the noise?
It's an interesting challenge from an engineering point of view because it's such a huge dataset, and from an analysis point of view because it's natural language and the structure is different depending on the source - tweets vs forums vs news site content etc.. So how do you bring it all together to give users the ability to get insights from this vast data set? Small innovations and significant but incremental improvements are happening all the time, but there are then the longer-term engineering breakthroughs which - if they can really shift the ability to analyse things at scale and make it easier to wrestle with - are doing a massive service for users. So we're trying to productise some of the big engineering programmes we've been on - they may have involved only 3 or 4 people but they're big efforts that have gone on for multiple years.
Secondly our application has been a bit like a Model-T Ford, you can use it for lots of different things as above, but it's the same model, so we've been pulling that apart into a jobs-to-be-done or use case-based framework; to take what we've built and think of it more as a platform to create applications that are easy to use.
Thirdly, we're learning about scaling the company and doing so globally - my leadership team is ten people, I think 3 in Boston, 1 in New York, 4 or 5 in the UK, 1 in France, and teams for other functions are also distributed, and we need to structure them so it's efficient but not siloed. We're learning as we go. Covid-19 has actually helped us work more cross-functionally, which is interesting.
Fourthly and finally, there's something we announced about a year ago when we acquired Qriously - a mobile-based survey company - we're trying to bring together different data sets to give a bigger picture of consumer behaviour and we're calling it 'Digital Consumer Intelligence' - so we're trying to expand the way we think of ourselves from social media analysis to this - taking things from different sources and trying to mash them up and make sense of them. In addition to survey app data from Qriously this includes things like online transactions, search data. Some of those are at a very early stage but I'm trying to think long term.
JB: Companies can no longer just leverage primary data - they now have to lay other data on top of that. It's really difficult to do... but even more relevant now in the age of Covid.
GP: Yes, there's a fundamental split depending on how the data is used within big companies: whether it's in a 1, 2 or 3-year cycle, long-term strategic input, or whether they're asking 'What's happening today and what are our competitors doing' and it's more tactical - with other data informing that tactical feedback loop, such as sales data or performance or online marketing and search. So the boundaries around the timeline of decision-making are really important. [If you're combining sources / types...] Having the skills very close by is important - maybe a tight partnership with an agency that can run quantifiable analysis quickly on data that's been gathered, and say whether they've found anything and what - we're seeing that 'market research data' feedback loop shortened dramatically inside a lot of our customers as they look to be consumer fit and adaptable / fast-changing.
Many of our customers are getting their budgets reduced but their workload increased, and the research industry is evolving from qual to quant, from quarterly to daily, from reports to quant data - or quantifiable insights rather. 'We've seen x number of people do this, we think this is a trend'. All powered by this availability of big data and the need to work with it.
... so it's an extra-interesting time in the world of MR, and it would to be honest be great to change the name of the industry which has a bit of a stodgy old school ring to it. By contrast, what's happening with consumer insights in some of the world's best companies today is really quite groundbreaking and innovative, and very forward-looking, so I think we're seeing a sea change in the MR industry. It [the industry] will get bigger, but it will change the way it is.
Getting started and making money
JB: I want to dive into the story of Brandwatch. You started the business in 2007, Facebook did its IPO I believe in 2012 and Twitter the following year, so you were really ahead of the curve.
GP: yeah so we should have got a lot further than we have to be honest [laughter]
JB: yeah there is that [also chuckling]. No but... then we hit a recession, directly after you started the business - how did you navigate that?
GP: I had been running this tech services business with some friends and I didn't want to do that any more, I wanted to build a product company. I started with not much idea, but I wanted to build a product. We had been working on this small project around building a web crawler, which was never really finished - and I used that as an impetus to create Brandwatch, I said to my partner I want to take this and he said 'you can buy me out and off you go'. He went and joined Google and latterly DeepMind, so an amazing guy... but I wanted to build a product.
We took maybe a year to build out the web crawler and then work on some basic sentiment analysis on what we'd found; launched it in August 2007 as an SaaS product, and our focus was on special interest forums - a UK one called moneysavingexpert - bank accounts and financial decisions, mortgages and so on. So we were crawling that site and one of the big banks in the UK wanted to launch a new bank account for young people, so they wanted to use our system to know how that was resonating inside that community. That was our first subscriber in September 2007 - we didn't get our 2nd customer until Jan 2008, which was a challenging time, sure! Trying to raise money as well, to keep the lights on.
Then the agency world in London - creative and ad agencies are always very strong here - started to realise there was this new ability to tap into consumer insights and understand them, so they started subscribing to it, not paying a lot of money. A couple of thousand dollars a month - not a small amount but I mean you need quite a few of those to keep a team of ten going! - and we started to see some traction, so we built the product with their needs in mind, and they had all sorts of different needs so that's why the product ended up being this very flexible system. In terms of the recession, it didn't really impact us.
Well, who knows because I don't have a control group, non-recession!...
JB: [laughs] that's fair!
GP: ...but we did pretty well throughout the recession mainly because it wasn't a crazy expensive product compared to people-based MR - not that we were really thinking of ourselves as MR at that time - but also even during the recession, the world was going online - social media etc.. smartphones, Google, Twitter - online was still growing like crazy and that was our domain. The move of money from off-line to online is the trend of the last 20 years, or one of them, and that's our playground so we were insulated from the recession.
JB: With an SaaS business, you really appreciate having guaranteed revenue in a recession...
GP: right - it's not guaranteed though!
JB: well OK...
GP: ...but yeah, the predictability of Brandwatch has been pretty good - except for the last 18 months, we've merged with Crimson Hexagon, that's been a much more unpredictable time, but up until that transaction we would put out a number at the beginning of the year and we'd be there or thereabouts at the end of it - we were always a little bit optimistic, but we'd be within 5 or 10% of it every year, because it's reasonably easy to look at the waterfall of money, sign up a customer, have a certain % of them renew, some of them grow, some of them churn. It's not a complex model.
The 5x Future
JB: are you ready for some rapid fire questions? Good. How do you think the MR space will be different in the next 5 years?
GP: more quant, faster, easier to use, cheaper tools, 5x bigger as an industry.
JB: [laughing / resisting the urge to immediately probe that large multiplier] Er... biggest issue facing today's market researchers?
GP: complexity of data, the amount of data, the time it takes to do stuff and the cost.
JB: and the three characteristics of an all-star employee?
GP: they're all behavioural, all about attitude, so:
1. willingness to learn; 2. ability to collaborate; and 3. work ethic.
JB: [laughing] Really great speed there. I want to delve into the phrase '5x bigger'. The space overall has been relatively flat at an aggregate level, according to ESOMAR for example. Why so optimistic?
GP: it's a nomenclature issue - there are certain things that go on today that are not called MR and are not captured, so it's the way it needs to be measured going forwards, but I'm optimistic because I remember an Uber investor speaking at a conference saying that Uber '5x'd the New York taxi market', not because more people wanted to travel, but because they made it cheaper, faster to use: use the app, it comes to you, you don't have to tip and that annoying stuff; but they were also cheaper. He said that 5 or 7 years ago but that stuck with me. Usage will go up - that's what's going on with MR - it's no longer just 'Find a focus group... write a report, deliver it 3 months after it was commissioned' - that doesn't happen much any more, so where we are making it faster and cheaper I think you will see usage going up.
Where you will see that on the client side... the companies that are able to adapt more quickly to the changing consumer landscape are more successful. If you're a company and you're not baking consumer insight into everything you're doing as an organisation, then you're likely to be out-innovated by someone who is: you're not adapting as quickly as your competition. If we can make it faster, easier and cheaper then it [what we do as an industry] will happen more - so 5x could be a massive understatement. It's no longer MR - it's decision management, and if you're not baking consumer insights into your decision management you are probably not going to win.
GP: I don't have one but I want to get one [laughs]... The cultural foundation upon which Brandwatch is built, from the very beginning - also Runtime Collective - is 'treat other people as you would like them to treat you'. So if I lived by that golden rule more than I probably do now, that would be good.
Other 21 YEARS / Mobile Age content now available (read and download the full supplement here):
Accuracy and Authenticity - How Smartphones Revolutionised Research, by Richard Preedy, Verve
Mobile Election Polling - Abraham Muller, CTO at Brandwatch Qriously, looks at how mobile methods can make election polling easier to implement and more accurate.
Delivering Great Mobile Qualitative: Four Keys to Success - by Ross McLean, Vice President, Mobile Qualitative at 20|20 Research
7 Steps to a Better Mobile Survey - QuenchTec Survey Designer Dennis Sewberath highlights seven ways to turn a survey into a 'beautiful conversation', for the sake of better feedback and higher completion rates.
You've Missed the Boat - RealityMine CEO Chris Havemann looks at the continuing opportunity for market researchers in a world of consumers and media changed almost out of recognition by the mobile revolution. Part 1 of 2.
...but there's another one along in a minute - in part 2, RealityMine CEO Chris Havemann looks at Why MR firms are 'fundamentally well-positioned', and where the immediate opportunities lie.
3 Mobile Marketing Myths - Andy Chandler, General Manager, UK & Ireland at mobile measurement and fraud prevention specialist Adjust, debunks three fashionable but misplaced ideas from this booming market.
Little Did He Know - Mario Paic, Global Head of Data Science, looks at the rise of the smartphone and discusses its central role today at Ipsos Audience Measurement.
Did You Get the Message? Reach3 Insights CEO & founder Matt Kleinschmit looks at the rise of the mobile messaging-based Survey - with case studies from work for Jagermeister, Hyundai and Kimberly-Clark.
An interview with Tugce Bulut, the inspirational founder and CEO of app-based feedback business Streetbees.