Using big data to engage and acquire new customers By

big-data

What with everything being stored in "the cloud" these days, there is actually a scary amount of our personal data stored in large databases.

Take Facebook for example: You enter lots of profile data, like your date of birth, age, gender etc. And then write comments like "look at my new iPhone" or "We're engaged!".

This information can quickly build up a profile about you. Everytime you 'like' something; we know you like it.
When you congratulate your football team's win; we know who you support;
When you post a photo of your new puppy; we know you're going to have vets bills.

Yes, we know.

Big data is everywhere. Now much of this data, you need permission to access. But some you don't.

Facebook: Assuming someone's profile is private, you would need their permission to acces their data. Sounds hard right? Who would do that? Everytime you "log in with Facebook" - you can be giving access to your data too. It's surprising easy.

Twitter: The same applies to twitter profiles - however, much of what is said on twitter is public and accessable to anyone "my new puppy"

Imagine if we could utilise big data like the pros and learn about our customers. Better yet, why not learn about new prospects? What if you could engage with them? What if they became a new customer?

The principle

The principle is simple: when someone mentions "my new puppy" on twitter, you could engage with them with a tweet. Start a conversation. Later, once you have built up a rapor with them, you could introduce your pet insurance product to them.

Here is a more complex example using Facebook. When you sell one of your existing products to a customer, you could get them to connect to Facebook - from that, you could prefil a lot of their insurance proposal from their profile. Cool, right?

Guess what? It gets better. Now you have access to their profile, you also have access to their friends on Facebook. There is a good chance that like minded people, with similar lifestyles, would be interested in the same insurance products.

Another thing you could do, to capitalise on Facebook friends, is to get someone to say they just bought your insurance. Maybe you can give a discount as an incentive. So now they're telling their like-minded friends about you! That's an endorsement! I'm sold.

Implementation

There are two ways (and every combination in between) to utilise big data. Pay for it with advertising or do it yourself!

Advertising - the easy way

Let's be clear. This approach is nothing new. Advertisors have been making use of your data for years.

They follow you around the web, track what websites you look at, what you search and what you buy.

It's no coincidence that when you're looking for a new conservatory, you start to see conservatory adverts all over the internet.

Both Facebook and Twitter have their own advertising networks. Twitter "promotes" tweets and Facebook intermingles adverts within your own feed.

They both use big data analysis to decide who is best suited to receive the adverts and a what time of day.

You can really drill down to the type of person you want to target with these adverts. Looking for a woman, betweeen 30-35, employed, who has just got engaged, but has also actually set a date for the wedding? Great - now you can target your wedding/event insurance product.

DIY - the hard way

This is harder, that's for sure. But you can be sure of much more bespoke integrations, far beyond simple advertising.

The connecting to Facebook example above, to fill out the proposal form, is exactly the type of thing that would need to be custom built.

Monitoring the twitter public timeline for key phrases and sending out an introductory tweet is free (beyond the inital development).

Them: "booked my holiday!"
You: "Don't forget your travel insurance! - linktoyourproduct #loveholidays"

It is hard however, to distinguish positive/negative meaning from this textual data. Natural Language Processing is often needed over simple text search capabiliies. On the other hand using "likes" or "retweets" is a superb way of determining interests.

There are of course, other sources of big data other than Facebook and Twitter.

Location based services like foursquare or networks that track locations of photos like instagram could point to a certain product need. "Checking in" at Battersea Dogs home is a surefire way of identifying a new animal in the home.

One you can do right now

You don't have to use external data to utilise big data - sometimes you have your own!

It is quite likely you have the date of birth for all your customers. Why not send them a birthday card?

Now, assuming many people receive gifts for their birthday... why not, in their birthday card, give them a 10% discount on your "Gadget Insurance" product?

It's worth a try right? It's easy. And free, if you use email!

Conclusion

Using big data sounds scary and complicated. But mostly the hardest part is identifying the people to target.

You should have a "persona" identified - an imaginary person who would be perfect for your product.

Once you have your ideal candidate, finding the data can be fun - especially coming up with new inventive ways to find and engage with them.

Ask us - we know of many automatic services out there and ways to target new clients. There are lots of tricks on the web - sometimes you just need to know how to combine them in the right order!

And if all else fails, we may even be able to develop you a custom solution.

Utilising big data takes big ideas.

By
comments powered by Disqus