Great things come in small packages. Three things we’ve learned about data that might seem surprising.


In a world awash with information and the concept of Big Data enjoying its moment in the sun, it might seem odd to be championing a less is more approach.     However, at Houston We Have, that’s exactly what we’ve been doing – with some startling results.

When clients come to us and ask for help in navigating the road ahead, more often than not they’re looking for us to make sense of masses of information so that decisions around how to proceed can be made more confidently.   A bit like shining a bright light into a darkly lit tunnel – our work can help our clients better see what lies ahead, so they can plan accordingly.   From early indicator tools, through to assessment models (with more transparency and less bias), we combine information with expert observation to generate plausible conclusions that can be used as a basis for action.

The thing that often surprises people when we get stuck into a particular problem is that large and complete data sets aren’t always as important as you’d think.   Here are three lessons we’ve learned that might not be what you’d expect from a company that specialises in Augmented Intelligence.


  1. Big isn’t always beautiful

More data isn’t always the answer.   In fact, it’s often more unhelpful than anything else.  Imagine trying to figure out if there’s a link between a frequent traveller’s propensity to fly more as prices drop.   Now imagine you’ve got access to all of that traveller’s flying habits.    One can certainly see how information on the trips they’ve made, and at what prices, would be useful.  Rather less useful might be information on whether they choose a low-sodium or vegetarian meal onboard.    For clients who worry that they don’t have loads of information, we often find ourselves reassuring them that this might not be such a problem.


  1. Not knowing something isn’t always a problem

Incomplete data has long been the nemesis of a thorough analyst.  Today, however, it need not be.   Imputation and inference are techniques available to data scientists and the clever folks at Houston We Have that allow us to derive information that might otherwise be missing.   The patented algorithms that power our Intelfuze software apply inference in a way that other tools are simply not able.   In real life, what this means is we are now pretty good at “filling in the gaps” where information might be missing from the original source.


  1. Nimble is key

This is probably the most important lesson we’ve learned at Houston We Have.    Clients (especially in larger more complex organisations) often tell us that getting all the moving parts assembled for a major new project can be overwhelming.   As a result, many great ideas die before they ever get off the ground.   We’ve found that by using the data we do have (rather than focussing on what’s missing), using inference to produce a more useful dataset, and creating predictive models that are designed to accept NEW information as it becomes available means we can produce assessments and pathways for action more quickly than initially thought.   Clients who initially thought only Goliath could provide the answers were pleasantly surprised to hear that David was actually more what they needed.

These three lessons have been learned from a number of engagements across sectors and applications.     From FinTech thought to Health, Defence and Consumer we have found that powerful decisions can be made with less information than you might think.    Talk to us to learn more.


For information:


Peter Hong
Country Manager NZ, GM Public Sector ANZ
+64 210 244 9784

Share on email
Share on facebook
Share on linkedin
Share on twitter