The most holistic source of client insight that every manager has, and no one is looking at

Do you want to tell what impact has your latest marketing campaign has had before clients start buying??Do you want to know what funds and strategies clients are engaging with you most on (and not just what you are trying to sell to them)??How about what new regulations clients are asking about??How has the latest external event affected client behaviours??Which clients are being over- or under-served??Want to know if your digital transformation strategy is working??Or how long it takes to answer a client query?

The data that can answer this huge variety of questions, plus thousands more, is sitting on one of your servers right now.?You just need to get at it.?And it doesn’t take huge systems implementations, expensive data feeds or armies of people to transform that data into real insight that can help you run your business more effectively.

What server is that??It’s your email server...

================================================

The vast majority of any asset manager’s client communication is done via email.?In servicing or operational teams, it can be close to 100%.?And in most cases no one other than the recipient sees the email.?It’s a transactional exchange between client and the recipient and once it’s concluded, they move on.?There is rarely any insight, data or analytics delivered.?

And it’s not surprising, really.?When we’re all pretty stretched anyway, who has time to log every email on a CRM or operational system, let alone tag it with metadata that will drive management information?? Unless it’s a “high value” email – a win, a performance update, a change to their account – or the recipient is incentivised to log it, emails are only ever recorded for posterity, to be re-read at a later date by someone who may be interested.

What we don’t get from that is any broader strategic insight into client activity as a whole.?

But the technology exists today that can deliver that insight automatically, without the overhead of people logging everything or drawing information out of emails.?Big data tools, combined with some basic data science, mean relevant information can be extracted from the thousands of client emails a manager will receive every year.?Combine that with CRM data to provide the client context, and a visualisation layer over the top of the data and suddenly you have the broadest insight imaginable into everything that’s going on with your clients.

Heads of sales can see every fund and asset class that’s being talked about, supplementing the higher value conversations that will already be recorded in your CRM system.?Marketing leaders can see the follow-on impact of their campaigns beyond the click-through tracking tools.?Heads of client service can see query types, resolution rates, bottlenecks and issues, alongside new trends and pointers to where to create solutions ahead of time.?Operational heads can see volumes and spot potential for automation.?And on and on.

Your emails are the best source of client insight and competitive advantage managers have, yet very few are getting proper insight from them.

================================================

My own journey with building this kind of insight tool started a couple of years ago.?I was running client service teams across different geographies and segments and was trying to understand the key metrics of the service organisation as we were under pressure to reduce headcount.?I knew client metrics – number, AUM, client type, products etc. – but had little quantitative date about the service we were delivering, other than what was captured on our account opening and transactions systems.?And I knew how my team were feeling – busy, overstretched, under-supported – but was struggling to win the argument that we were busier than ever despite client numbers falling.?

So we embarked on a journey to capture client query volumes and types.?We built a plug-in to Salesforce, our CRM system, and tried to make it as easy as possible to capture client queries that would drive the needed management information.?But it was a disaster.?The team resented having to log queries despite the process being as quick as we could make it, other teams felt the operational data was “clogging up” Salesforce, I was getting grumpy because the data being delivered was so inconsistent it wouldn’t stand up to rigorous challenge.? We ended up canning it and I had to look for other ways of getting the same information.

A random conversation with our Head of Audit, of all people, led me to a new team of data scientists who had the kind of tools I was looking for and were enthusiastic about the challenge.?We designed a data feed from our CRM system that gave us clients, contacts, segmentation information, geography and AUM.?We used that to extract client emails (leaving behind internal emails) from a selection of client service and operational mailboxes, with some clever code that filtered out a lot of the noise that’s on an email trail, like repeated footers and disclosures.?At the same time, we built a natural language taxonomy that we would use to interrogate emails.?So, for example, when looking at query types, we said, “if the email contains words like statement or valuation, call them all valuations and bucket them under operational query.?If the email contains words like carbon, voting, climate, call them ESG” etc.?

We then combined the data sets with the taxonomy and developed a visualisation layer that could present large data sets in an intuitive way, as well as sort and filter the information.?

The results were extraordinary.?

For the first time I could see client email traffic across the globe.?I could see how that correlated with client value and client type.?I could see what types of clients were asking what questions to which of my teams.?I could even see the interactions with consultants and intermediaries that have such a big influence on the team’s work volumes.?Even though we were in a “pilot” with only a handful of mailboxes hooked up we could see patterns of information that we had never seen before.

And the tool was really flexible.?Throw in queries containing the words ‘Ukraine’ and ‘Russia’ and we instantly saw the huge spike of traffic early last year.?But interestingly the client queries calmed down long before our own internal discussions and response, showing it’s easier to respond to a crisis than it is to return to business as usual.?We also experimented with tools that analysed sentiment in the language, which showed a lot of potential, though we only scratched the surface.

The flexibility and depth of the data set meant it could be used to answer pretty much any question and, with the modern technology that was at our disposal, do so virtually real-time.?It was an incredibly powerful tool.

There were a few key things I learned along the way.?First, it’s important to set the right culture and goals at the start.?People can be sensitive when you say you’re going to analyse their emails so it’s important to be clear that the only emails being analysed are to or from clients, not internal email traffic.?Second, the goal is to see patterns of client behaviour and to draw out insight and measurements that help run the business.?It’s absolutely not about reading people’s emails or being ‘big brother’.?You can support this goal – as well as the culture – by building a rule set that restricts access to the underlying emails appropriately, allows flexibility around sensitive clients and, of course, takes into account national data residency and privacy rules. ?Finally, it’s helpful to set a limited number of key strategic questions you want to answer up front.?These can be used as an initial goal that will allow you to fine-tune the system and prove out the process, as well as gain leadership buy-in.?Once you are there, you can then open it up to broader analysis, but it has so much potential that it’s easy to get lost down a rabbit hole trying to answer every question that comes up on the fly, which can derail the initial delivery.

All this insight is sitting there for any asset manager to see, if they commit to taking a data science approach to email analysis.?It’s child’s play to many other industries, but something that very few asset managers have ventured into.?I thoroughly recommend it!