I got a demo of the Talent Analytics platform recently. Talent Analytics is a company that develops solutions that provide relevant information to help companies anticipate the impact talent has on their business goals. Advisor, their platform, is a web-based SaaS platform.
At the top level the product allows one or more organizations to be managed – many users of the platform are management consulting firms and they typically have multiple clients. Talent data is collected from each company with data being managed around the people in that company, grouped by role or functional area. The people being managed might be employees or candidates. Data is gathered about these people through a 61 question survey that takes a little less than 25 minutes to complete. The questions were built based on psychometrics and other research and are periodically validated statistically. The questions are randomly displayed and responders rank the 4 answers presented from most to least appropriate. The results are processed using the various algorithms that Talent Analytics have developed. Questions cross-check each other and are designed to catch both people trying to game the survey and those who are just randomly picking answers.
For each person the Talent Analytics algorithms create a series of analytic scores or metrics. For instance they have a set of scores for the way people think about business identified with as DISC –Decisive (directive v collaborative), Interactive (social v reserved), Stabilizing (likes predictability v likes change)and Cautious (rule follower v informal). These scores give a quick overview of people’s performance style – how they would like to work. They also calculate a set of Performance Styles values for how people feel they are performing in their current role (people often behave differently than their preferred style perhaps because they believe they need a certain performance style to be successful). Presenting this data in a simple table immediately shows interesting things about a selected group of people: A large gap between preferred and actual style is likely to cause stress, a team member with a score that is very different from the average of a group might not be a good fit for that team and so on.
They also measure ambitions on 7 different dimensions. This shows the biggest drivers for people in their own assessment of themselves. Each person ends up with a ranked list of the 7 dimensions showing the relative importance to them of Economics or results; Altruism; Theoretical or love to learn/teach; Authoritative or mission focused; Political or competitive; Individual or non-conformist; and Creative who like being out of the box. These too are presented in tabular form and again the differences in attitude across a team, for instance, can be telling.
Both the 4 Performance Styles and 7 ambitions data are designed to be stable – they are things that cannot easily be changed or trained for. The four Performance Style parameters are genuinely independent variables – Talent Analytics have done extensive analysis to make sure that values of one don’t predict the others.
Besides the simple tables showing a group or team, this data can be presented in various graphical formats. Showing it on a graph can show the differences between team members, between candidates or between existing teams and candidates for the team. This can be used to compare candidates, to see how a team clusters, to see how top or bottom performers look (they have found that top and bottom performers often show particular patterns within a group or role) and so on. It can also be used to compare people from multiple teams that need to work together or even compare teams in different companies to see if their relative performance can be explained.
The tool also allows the definition of benchmarks. There represent role requirements in terms of these various dimensions. People can then be compared with these benchmarks to see how well matched they are. You could therefore see, for instance, which sales reps have a style that is the best match to the benchmark for a sales manager rather than simply promoting the sales rep with the best results (who might be terrible fit). The tool explains gaps between a person and the benchmark and these explanations can include custom questions to draw out why the gap might be there (for use in hiring) or what might be doable to help manage the gap with training or mentoring (helping someone tone down their aggressive style say). Of course people who show a big gap against one benchmark could be compared to other benchmarks in the organization to see where they might fit better too, allowing for coherent reassignment of people when they don’t fit in their current role. Talent Analytics’ partners often focus in a specific vertical giving them unique insights into what is an appropriate benchmark for a role in that vertical.
Talent Analytics have found that, while the tool does not change people, comparing teams and people this way helps give them concrete things to discuss. This helps groups or individuals who must collaborate but struggle to do so to get out of name calling and complaining and into constructive discussions of how to manage the difference in perspectives. In this vein the tool also offers an overall dashboard with dials showing the overall pattern of people in a company or a division or a team. The dials show the range of values within which groups fall across the company, allowing the differences between groups to be clearly seen. This could also be used to compare the approach of different groups that would have to work together if a proposed merger was to go through, helping with planning and perhaps even valuation.
One neat feature that is coming soon is what they call an Enterprise CUEcard™. The idea is to provide a Talent Analytics vCard that can be merged with a typical company address book profile. The CUEcard contains advice about how to interact with each person (“Don’t patronize him with incentives”, “Don’t be sloppy” etc). The tool will also be able to generate a Team Playbook™ – a PDF with everyone listed, overviews of the Performance Styles distribution of people in the playbook and tips and traps for communicating with each person. Again these might be all employees, a multi-company team in a Joint Venture or a new team that will be formed out of a merger.
To date Talent Analytics have mostly worked with companies in the 1,000-8000 employee range but the size of customer is growing and they have, for instance, a bank using Advisor for call center representatives. This bank had 100% turnover, often losing people during the training, and this meant they were hiring and training (and losing) 30,000 people a year. Some companies just use it on their executive team and leadership and many others start there and work down the organization. Talent Analytics continually investigate new statistical techniques and are working on questionnaires that are more dynamic, that learn from results and ask different questions as they go.
One feature I particularly liked is that the analytic scores are easily extractable and can be integrated and used alongside other data. This would allow, for instance, predictive analytics that used Advisor data and other HR or performance data. More on this in a future blog post.