Skip to main content

Interview: Why personalisation is key to leveraging big data

Big data. No doubt a phrase you've heard countless times before as more and more companies advocate the use of data collection with the aim of better serving their customers.

But, not all companies are making the most of big data. To shed some more light on the matter, we spoke to Neil Capel, founder of Sailthru, to discuss personalisation, predictive analytics and the key industry trends.

  1. To start off, give us an overview of Sailthru’s journey to where it is now

When I founded Sailthru in 2008 it was with the vision in mind that every customer is unique and that marketers needed an easy solution to deliver on that idea. Fast forward six years and we’re 170 people with more than 400 customers across multiple continents.

You never know how things are going to evolve and the Sailthru team all shares my excitement, passion and desire to continue to grow. What’s been most significant about our journey from my perspective as founder is how marketers respond when they get their hands on the technology.

Some come to us just looking to send better emails and we not only deliver in spades (175 per cent increase in revenue per send), but their ability to grow with us is what ends up making the difference. It’s easy to implement for onsite, easy to connect to mobile and easy to connect to social. The ability to automate decisions across multiple channels from a single data source is promised by many, but we’ve been natively building that capability so that we’re not just another frankenstack.

That’s always been what we’re about and it’s critical to where we are going simply because that is what marketers need most.

  1. From the companies you talk to, what is wrong with regards to their approach to big data?

The most significant challenge is that many companies don’t approach data understanding both where they want to go with it and where they are today with it. For a company to leverage big data effectively they must be goal oriented: do you want to increase customer lifetime value? Do you want to increase total retention rate?

By having an end goal in mind you can develop a strategy that can deliver on this. Equally important is understanding what your data blueprint looks like today: what data is collected? Where and how is it housed? Who is the owner? These are just a few of the questions that need to be asked.

Most companies go right into the middle and work to find solutions to their big data problem, but when it’s not flanked by the context of where they are today and where they want to go the approach fails to deliver.

  1. How can big data be leveraged across devices?

Big Data can be leveraged across device by attributing data to your individual customers, understanding who connects where, how they use those devices and how they want to engage with you on each device. It’s critical to understand this at the individual level because we are all so incredibly different in our behaviours (save for waking up and going right to our phones), and it’s critical to cross-channel engagement: unless your customer experience is founded on a single view of the customer, you won’t be able to manage real-time shifts in your marketing as they shift - in real time - from device to device.

Brands must be sure that the data they are using to increase the relevancy of the experience across devices and channels is a pool of owned, deterministic first-party data, rather than inferred third-party data. By using third-party data in hopes of augmenting the customer experience, most brands will find this approach backfiring as it’s not data that the customer knows is being shared.

A great example of this is one publisher that we work with. They’re one of the fastest growing business titles in the world. When I’m on my way to work and read an email from them on my mobile and click through to consume the content they know where I came from and what I’ve just read. When I’m then on my desktop or tablet just a few minutes later what I clicked through on email earlier that morning doesn’t appear in my stream.

This experience is a large part of what keeps me reading and it’s why some of the most well known publications in the world are struggling to keep up with the new cohort of publishers who get that in order to remain revenue focused the quality of the reader experience is just as important as the quality of the content itself.

  1. Just how important is personalisation and what can companies do to improve this aspect of their business?

For me the question is less how important is personalisation and more ‘how important is it to attract better customers, deliver a modern experience, create loyalty and increase customer lifetime value?’ Many view personalisation as a campaign tactic, but the truth is that it’s a business strategy. It’s not just onsite recommendations, it’s about creating an contiguous experience for every individual that increases in relevancy with every action that they take; it’s about making every individual channel more effective by using data from all channels.

Gartner tells us that organisations that fully invest in personalisation by 2018 will outsell their customers by 20 per cent. That’s huge and when coupled with what we know from our own data - that 80 per cent of a brand’s future revenue will come from just 1 per cent to 10 per cent of existing customers - it makes personalisation mission critical.

What companies can do to improve this aspect of their business is to understand that personalisation doesn’t have an end point. It’s about always adjusting to the individual user. It’s a methodology that has a maturity curve. The goal is to extract the highest value from the investment!


  1. Why should companies be focusing on predictive analytics?

Being able to predict what each individual customer will do next is the holy grail for marketers. At the core, behavioral predictions can enable brand marketers to predict customer lifetime value so that they can identify their most valuable customers and then work to retain them while at the same time acquire customers who share their attributes. For channel marketers, behavioural predictions enable the ability to significantly advance segmentation from the individual level up based on their future behaviours rather than on backwards-looking reporting.

One of our clients - Country Outfitter - a US based lifestyle brand uses Sailthru’s behavioural predictions tool to identify their best customers based on the likelihood to make a purchase in the next seven days. We’re making that prediction for Country Outfitter for every single one of their customers on a daily basis so that they can then group those individuals into a top customer segment and ensure that they are engaging with them appropriately.

Using this approach they were able to increase the size of their top customers segment by 70 per cent (they were previously using a form of RFM modeling) and they saw a 72 per cent increase in email revenue. They then expanded their approach to optimise acquisition and in one test they saw a 43x return on ad spend on the campaign based on predictions vs a 4x ROAS on the campaign driven by reporting.

Predictions can revolutionise how marketing teams operate, but it’s not easy to determine who can truly deliver. Many of the technologies promising predictions are actually just providing onsite product recommendations. It’s essentially rebranding something that has existed for years. Instead of saying ‘if we recommend content/products that this user is likely to be interested in they will convert’ they’re now saying ‘we predict that if we recommend…’ Or they continue to operate at the broad segment level, instead of getting down to the individual level, and continue to use fairly rudimentary approaches such as RFM modeling instead of algorithms that learn from a large volume of data on each individual.

To choose the right approach marketers must first educate themselves on the nuances, identify what challenges they are looking to solve and then the partners who can support them in that process.

  1. What are the main security issues companies are coming up against?

Right now there is a wave of DMPs selling themselves as the answer to a fully connected, contiguous customer experience; however, the approach is plagued by the challenges presented by third-party data. DMPs work best when used as a single component of the acquisition tech stack; however, when used to manage relationships and retain customers they present a significant challenge for marketers and technologists alike in customer privacy.

The digital media sites that I read every morning shouldn’t know what I purchased online the night before. And that’s just one example assuming that the data being used is both recent and accurate, which you cannot guarantee when it’s third-party. By focusing on solutions based on first-party data collection organisations can work to build the foundation they need to manage customer privacy.

  1. What tips would you offer to smaller companies looking to make the most out of their customer data?

Start collecting the data now. What’s most important is that you have the capability to collect meaningful data and attribute it to your individual customers.

By starting to build this long-term customer data asset today, you’ll be able to quickly transform your marketing-based customer experience when you have the resources to do so. You must be looking at far more than just opens and clicks and treating the rest as digital exhaust: every behaviour has value!

  1. How has the approach to big data changed in the last 12 months and what trends do you expect to see in the future?

From a marketing standpoint, big data continues to be a significant challenge. The enterprise marketing clouds that have been built through acquisition (without progress on integrations) are not enabling an actionable single view of the customer, which has stalled progress in translating meaningful data into insights-based action. So the output doesn’t end up meeting the promise.

As marketers begin to understand that before solidifying a partnership that they must have a roadmap for how they want their customer experience to evolve and understanding what requirements will enable that, we will continue to see companies stuck at the impasse.

This is what I believe will drive what we see in the future: marketers choosing solutions that enable them to desilo, enable native cross-channel engagement in real time, and in the end, enable them to get ahead of their customers.