“You have to really argue to be against data. In a club, statistics give people the chance to exist who have little knowledge. And these numbers act as a safety guard for decision-makers who lack courage. That, yes, that annoys me.” - Gilles Grimandi, Arsenal chief scout (L’Equipe, 30th December 2014)
Does a football scout have a better understanding about the barriers towards data and analytics in organisations than many business managers? Based on the comments he made in an interview at the end of last year, Grimandi may well prove to be the most insightful commentator on the up and downsides of transforming around evidence-based decision making. (You can read a translation of the full article here: http://www.getfootballnewsfrance.com/2014/giles-grimandi-statistics-enable-people-with-little-knowledge-to-exist/) Given the popularity among business conference delegates of hearing from sporting heroes, here are two lessons which sport can teach data and analytics practitioners:
1 - The only number that really matters is the final score
The arrival of Sabremetrics in American baseball had a profound effect on how clubs pick players and the way the game has been understood. Popularised by the book and film “Moneyball”, the use of statistics to assess performance and potential has since filtered out from this home turf and into a growing number of other sports, most recently Premiership football in the UK.
Numbers generated on player actions, from scores and assists through to distances run and touches of the ball, allow managers, coaches and owners to understand better how much value they are getting from their single biggest investment. Players showing highly positive numbers will be sought, while under-performers are quietly moved on.
All of that is useful for the activity that takes place within a football (or any other type of sports) club between actual matches. Used well, it should improve the outcome of those games. But at the end of the day, it is only that outcome - the final score - which really matters. Assembling a brilliant team with optimal statistical figures is only worth doing if it gels on the playing field and becomes more than just a bunch of names that have good scorecards. For example, Real Madrid has won La Liga five times since the advent of its “galacticos” transfer policy in 2000.
Business is the same - each project, campaign or activity has a specific end which should have a target and clearly defined metrics, such as making a set number of sales, achieving a productivity rate of x or hitting delivery promises n per cent of the time. (If such metrics are not in place, data and analytics may be having as much effect as the moon - nobody will be able to tell.)
The job of data practitioners and analysts is to support the drive towards that goal, not to generate ever-more beautiful or persuasive numbers.
2 - Risk can be quantified. But risks still have to be taken
Grimandi sees evidence-based decision making as lacking in courage. That may be true in the heat of a football match when rapid choices need to be made about changing players or formation. Few managers will have the time to call up a new spreadsheet before telling a substitute to put his boots on.
In many cases, the decisions to be made in business do not involve such high stakes. A lot of business activities are repetitive or low value, but they can be “tuned up” by applying analytics to understand where an incremental gain could be made. As Sky and Team GB discovered in cycling, add up all of the hundredths of a second that might be saved through minor tweaks and you end up coming down the home straight ahead of the field.
Even where a company might be facing a big gamble, such as a new product launch or pricing policy, data and analytics still have a part to play. Modelling scenarios and outcomes in advance can help to ensure that alternative options are developed and mitigation ready-to-hand, for example.
Knowing how much is at stake is vital to anybody in a sport. But once you have that number, it is ultimately a choice about whether to act on it or not. As any statistician will tell you, the best model in the world is still never 100 per cent accurate and outcomes may fall below (or above) the predicted data point.
Leadership in business is no different from management in sport - it takes guts to make the call.