Data Mining and Modelling
At Stillae, we simply ‘do not do data in isolation’
Our first concern is to understand the business context, strategy and stakeholder priorities, as these are critical –they are the reason for our work and provide meaning to analyses and projections. We believe that great analytical work can count for nothing if it doesn’t speak to this context, if it lacks buy-in from stakeholders, as at the end of the day it is our clients’ decisions and actions that matter.
You have in-house analyst teams
We always work with your in-house people to integrate their tacit knowledge, build capabilities and transfer knowledge. They are stakeholders themselves, and are critical to embedding the work and implementing changes. We often find that in-house Finance, HR and IT teams are busy with on-going reporting duties with little time for, or experience of, exploratory analyses with data from multiple sources, which is so critical to uncovering genuinely new insights.
Where our clients have Specialist/Advanced Analytics Teams in-house, we partner with them, building on existing work with a fresh external perspective. They value our pragmatism and creative thinking, our wide experience with all sorts of data, and particularly our ability to secure appropriate executive attention.
Your existing data is imperfect –incomplete, inaccurate or insufficient
Data is never perfect, so in many ways the first question is always How bad is it? In our experience there is usually value locked away, despite the imperfections. We use statistical methods to explore how reliable the data are, and we then build in these margins as we develop our models. This allows us to develop estimates and calculate possible ranges. Its often possible to triangulate a problem –to narrow in to a close range of possible answers by taking several different approaches (none of which is accurate enough on its own). We’ve often found that this can get you close enough to the exact answer to enable confident decisions.
But of course we can’t work miracles, if the information simply isn’t there, then we can advise on what you should be collecting to close the gap.
Independent of IT infrastructure or software platforms
We don’t sell software or IT systems. We work with the data that you have, focusing on drawing value from it and helping you put it to use. The functionality of your system is relevant, in as far as it impacts how data is or can be used, so we offer advise on functional requirements and help clients to plan future developments. We very much like IT geeks, but we’re just Stats geeks and we stick to what we’re good at.
Our step-by-step approach keeps you in control
In our initial meeting we assess whether analysis is likely to yield fruit. If we don’t think it will, we’ll say so, as we don’t want to waste your time, or our own. We usually begin with a data audit. This allows us to establish more precisely what could be done. And we then work iteratively –identifying what analyses are possible (and the effort involved in each), allowing you to remain in control of where effort is focused.
Finding the right level of involvement of your internal resources
Clearly we need to understand your business context, and to understand what data you have and to access it. So some time will be needed. Beyond this though, we can involve your internal team as much or as little as we agree is appropriate. We’re often asked to build capability and transfer knowledge, but sometimes speed is of the essence, or its only useful to brief the wider team on findings (rather than involve them in exploratory forays). We’ll agree on the best approach with you, based on the specifics of the situation.