At this year's Europe's Customer Festival some of Big Data's big minds came together for a short fireside chat to share their ideas about how Big Data needs to be built into your company as core concept if it's going to help benefit your customers.
Marc Cregor, Regional Sales Manager of HP Vertica, moderated the panel, while former CTO of Obama For America, Harper Reed; Liberty's Divisional Director of Loyalty Strategy, Sonja Fourie; and Barclays' Director of Business Performance and Analytics, Dan Bolland, had the lengthy discussion over Big Data implementation.
It was really interesting to hear how all three really did agree that Big Data should be one of the most integral parts of any company. These people should be there, relaying the information right back to the top level. After all, they know best about what customers are actually doing. The consensus is also that just holding all this data isn't enough for companies. If anything, they should stop trying to get new information and just use what they have far more effectively.
As explained in Harper's talk about how Big Data helped the Obama for America Campaign, he genuinely thinks the term âBig Data' is "bullshit". Therefore, he's the first to suggest that this information companies hold needs to be used to empower users and improve their experiences. It's there to make life easier for customers and to answer their questions before they even ask them.
Dan chimes in with this too, saying that we need to better utilise external data sources and make use of social data too. Why rush out to get more data when you can't even use the information you already have?
Sonja, on the other hand, believes that customers should be focusing far more on taking your brand into the emotional space through the power of data. It's here that you'll be able to create loyalty with a customer and really make all that information you hold worthwhile as it can be acted upon quickly and relevantly – thus making a decent investment for you.
Of course, it's not easy to do. You can't just snap your fingers and have it happen, nor expect to get feedback quickly. So, how do you go and do all of these things?
Well, Harper offers some insight from his past experiences. Having come from an engineering background, the idea was to test, test and test again. If something didn't work, you knew pretty quickly and could just go and test it again. This can be exactly the same with data. Why not try something using your data and then, if it doesn't work, just try something else? Over time you'll gain a far more nuanced view of what your customers want and also stop tracking the useless data that just bloats your system.
Doing this will most certainly be a scary process, especially once you strip away the rubbish metrics and use the ones that really matter, but it will help in the long run.
This is something that Dan completely agrees with. He's seen KPI sheets that extend over two pages of A4, nobody needs to track that much – and if they do, perhaps they should re-evaluate how effective their business really is. You don't need to keep attaining data that you already know. You don't need to run with intuition. Now metrics exist, you know what works and what doesn't.
This doesn't mean don't use intuition, but use it in conjunction with data. If something feels like it might work, go and see if the data correlates with it. If it doesn't, then it clearly doesn't work. Harper even went so far to say that you shouldn't be setting out to do things that are "clever" or "cool", you set out do get useful information, and that's all you should be doing – and doing it well.
So, you can do this better by segmenting your data properly. It takes time, but proper segments help ensure you're running a tidy and tight operation that delivers the results you actually want to see. A lot of the time people get sent information that's not related to them at all. We should know if someone's bought something on Amazon as a gift – so related products shouldn't be showing up due to it being a one-off purchase. Take a look at what information you're getting and break it up into different segments so you can get information on a granular level – it'll help in the long run.
That's all well and good, but surely the real issue is getting useful data from consumers in the first place and getting them to act upon it?
Luckily, Harper, Sonja and Dan all agree that this isn't an issue at all.
In the past, many people have asked Dan about how to drive sales with big data, nobody has actually asked him "how do I do Big Data better?" But in reality, it's not difficult at all. You have to know what you want from it – i.e. what will help drive your sales – and then you need to keep your core message and push forward with that.
As Harper put it, about creating interfaces for you to use with customers. It's about creating opportunities to make people think and engage. People don't want to have these highly-focused emails, mailings etc. For many, it's an increasingly creepy feeling of being followed by a company they shop with.
Instead, create generic mailings, but specialise them with certain pages and offers others wouldn't get. This way it seems like a mailing to everyone, but those little tweaks make everything feel far more relevant.
People aren't dumb either, they all know that you're wanting to grab their information to sell things to them, there's really no point in trying to hide that fact at all, nor sugar-coat it, all that'll do is alienate people. And when it comes to the world of Millennials or GenY or whatever other absurd term you want to paste onto those plugged into the world of technology, they're even more aware than ever about sharing their data. Except, they'll do it if they feel it's worth their while. Over 25s may well want value or money in exchange for their data, whereas under 25s will want usability and functionality, they want something that benefits their every-day life – they aren't concerned about accumulating points to take money off an occasional shop.
They use Twitter, Facebook, Snapchat etc. because they get a service they deem useful out of sharing their data. You need to merchandise data acquisition to customers, you have to see value in the trade. It's something especially evident in the world of the startup, who grab more and more of this desirable customer information, and the reason why they do it is because they're using the same language as the consumer.
They offer something that they would want as consumers. They talk to them as if they're on the same page, and they keep the hell away from marketing terms and general nonsense words – using plain English to get their point across.
In essence, what Harper, Dan and Sonja conclude is that in reality nobody gives a damn about Big Data except for those involved in it. Companies just really want results and don't mind how it's done, and that's the completely wrong way to go about bringing data into your business.
Here's the tl;dr for you skim readers out there:
- Bring Big Data results, and those who deal with them, to the top level meetings
- Test, test, test. If something doesn't work, just try something else – you've got nothing to lose
- Use intuition and data together to find new avenues to explore, but don't ignore the cold hard facts
- Make sure you have segments for everything – with offline and online as an absolute minimum
- Create offers that engage with customers and make them think
- Produce a product that has worthwhile trade value to people offering up their data – otherwise they won't do it