our views and our knowledge in analytics and other releveant topics

our blogs

2018- The Year of the Insight Nuclear Submarine

NewsLetter-Submarine Jan18.jpg

On April 9th 1963, USS Thresher, the first of a new class of nuclear submarine for the US Navy, went on what is euphemistically called “eternal service” off Cape Cod in Massachusetts.  It was being put through its paces to understand its maximum operational depth when it sprang a leak and sank with all 129 hands.  The cause of the leak is generally regarded to be a silver alloy, used as a solder for some of the joints, not being able to cope with the pressure at 1,300 ft depth.

One of the problems of building a submarine is that it you can’t create a mould big enough to build it out of a single, continuous piece of metal.  That means that you have to create the pressure hull out of smaller sheets of metal, and then join them together.  And that means that you will have potential fault lines in your superstructure, where one piece of solid metal which will naturally withstand large amounts of pressure, meets another.  For most structures, even other types of ship, this doesn’t matter so much, as they will not be subjected to abnormal levels of pressure.  But for a submarine, that’s exactly what will happen; the deeper the submarine goes, the more pressure it will have to deal with.  As the sinking of the Threshershows, any failure under those pressures, however slight, will be catastrophic.  This is perhaps the greatest structural challenge in modern ship science, and one that was only starting to be understood in the early 1960s.

The modern solution is to embrace the problem and turn it into an asset.  What if the solder is constructed so that it is actually stronger than the surrounding sheet metal.  Using a mixture of very strong metals, and working them at very high temperatures, the soldering linking the metal becomes a powerful exoskeleton, rather than a weak point; the joints are in fact stronger than the surrounding plates.  Suddenly, the limitation in the submarine’s ability under pressure becomes the strength of the metal, not the joins around them.

I think it’s a useful analogy for where digital insight technology is at the moment, and the challenges that insight and digital directors face.  Previously, it made the most sense to buy a single marketing cloud and all its components (the equivalent of building a submarine from a mould of a single piece of metal), which meant you never had to worry about where the joins were.  But now, for the first time, the component “best of breed” tools in martech are developing faster than those from a single marketing cloud vendor, and so it makes the most sense to build your digital submarine out of the different pieces of metal, and sort out the joins yourself.  And this is changing best practice for delivering insight and the digital actions off the back of them.

The multichannel analytics technology marketplace is maturing, and fragmenting.  As analytics capabilities develop and grow more sophisticated, so more bespoke tools, or technologies designed for a particular purpose or part of the customer insight journey, are built and gain traction in the market.

Having been through an acquisitive phase in recent years, the main marketing cloud players (Adobe, Google, IBM, Salesforce) are busy integrating these businesses into their existing stack.  This means that their appetites for further acquisitions have been dimmed (at least for a while), so the smaller, more niche players are being left alone to grow.  In any case, there are now many smaller players in areas like attribution, or alternative digital data collection technologies, or tag management, that it would be difficult for the bigger players to accumulate them all.  And several of them are building standalone propositions that are either taking on areas of the big vendors directly (Tealium), or genuinely add new types of data not available in the big players (Decibel Insight) or are diametrically opposed to the approach of the bigger players (Snowplow, Celebrus).

This means that the multichannel insight and analytics technology space is at its most diverse for perhaps 10 years, at a time when expectations of seamless customer journeys are at their most sophisticated.  So, if you want the broadest range of analytics data, you will have a range of insight tools from different vendors.  And you may look at them independently, but to generate the greatest value from your investments, you will want to integrate them (or at the very least, their data) with each other.

And this is where the submarine metaphor comes in.  Because it can be difficult to integrate different tools.  So, when you are joining data from different tools, it’s very important to get the soldering right, but also more difficult.

So far, the difficulty of integration has been a big factor in what organisations choose to do.  Some platforms can’t consume large amounts of data from other sources, so you will need somewhere to store them.  That’s relatively inexpensive these days, as you can put it in the cloud, or your own data lake, but it does mean that you will put more emphasis on the joins between them.  And that part is up to you – your responsibility, your risk, your problem.

This has meant that many organisations have taken the easy option of buying more from a single stack solution; this is the equivalent of building your submarine from a single mould, or at least where the soldering is made out of the same metal as the rest.  Even if there are integration problems, at least then it is the marketing cloud vendor’s problem and you have one “throat to choke”.

However, we are starting to see more and more problems with integrations within a single marketing cloud.  This is the equivalent of taking your submarine out to test its performance under full operational pressure and having to abort because the soldering was made from the wrong material.  It’s crucial for the single marketing cloud providers that they get this right – they should be developing the very best joins for their solutions, otherwise, they undermine the value of buying into a stack in the first place.

However, so goes the new logic, if you are still having problems even with a single vendor, then you might as well take a best-of-breed approach, and take the risk of integrations on yourself.

Moreover, we are starting to see more “soldering” technologies appear, whose primary role or positioning is to join different marketing cloud tools to others, or to data lakes.  Tools like Syntasa have emerged to provide a framework solution to join data from different platforms together, or to your data lake.  This means you can make your technologies link correctly, and they can operate under pressure, without the risk of blowing a leak.

If you want to talk about how you can build the insight equivalent of a submarine with the right integrations, please get in touch with Station10.