Both the Obama campaign and the Romney campaign used data analytics heavily, but while the Obama team insourced it using Harper Read and a team of hired and focused analysts, Romney chose to outsource the effort, relying on the firms TargetPoint and Buxton to create the data acquisition process and analyze the result. Details of both parties’ efforts can be found on the technology-centric Slashdot Business Intelligence section here.
This isn’t a problem with the products — although it looks like the companies Romney chose are hiding from any connection to the election results — it is a problem of approach. If you outsource intelligence gathering and analysis, you’ll be at a strong disadvantage against someone who creates this capability internally and it would appear that the Democrats did more with inferior technologies than the Republicans did, showcasing this internal advantage.
Let’s talk about why it is unwise to outsource intelligence gathering this week.
Years ago, I worked in IBM Market and Business intelligence as an internal analyst during a trend when this was being outsourced. This was before Big Data analytics and much of the work we did through surveys, primary research (this was before the Internet) and purchased reports. The information, when properly applied, resulted in better products, happier customers and better competitive moves (tighter pricing, better branding and more targeted marketing).
If you know the enemy and know yourself, you need not fear the result of a hundred battles. If you know yourself but not the enemy, for every victory gains you will also suffer defeat. If you know neither the enemy nor yourself, you will succumb in every battle.
Reading Sun Tzu is refreshing because it is often easy to see the firms that are more likely to lose and those more likely to win just by applying Tzu’s stated strategies to the visible behavior of the firm. In this instance, it was clear that the Democrats had done the analysis, figured out where the Republicans were weak, set the strategy, and assigned resources so they couldn’t win. The Republicans thought they knew the Democrats’ weaknesses but clearly didn’t know their own and failed spectacularly.
When you insource intelligence, the employees who are working for you are more likely to recognize that a failure in intelligence or analysis can lead to their own failure along with the company. In addition, generally part of the staffing is done with people who know the market and the firm. When you outsource there is a far lower likelihood that the outsourced firm has any understanding of the market, the company and certainly not the competitors. It appears the firms Romney used are being outspoken (per the above link) on not understanding politics. Further, the outsourced firm wants happy clients and the happiest clients are those who get intelligence that agrees with their preconceptions of the world and how they would like it to be.
This lack of company, market and competitive analysis coupled with a high need to both show competence and have a happy client tends to result in bad data. If you ever wonder, for instance, why external number firms rarely show markets declining, it is because people don’t want to buy negative reports; they want positive reports they can use to justify funding.
An internal group, on the contrary, will be measured on its accuracy and is much more likely to provide a more accurate assessment or err on the conservative side. Over time, this is largely why internal intelligence groups have fallen off in favor of outsourcing and why firms often seem surprised that markets they are investing heavily in don’t grow as expected. Managers prefer positive reports and will fund them even if they are inaccurate. They will look for ways to pull funding from negative reports even if accurate (often waiting until a mistake is made before pulling the trigger).
In IBM, there were intelligence groups like my own that focused on accuracy and groups that focused more on outsourcing work and providing more positive results. The former groups were largely disbanded and the latter groups promoted more aggressively. Though, and I think this should be pointed out, both tactics led to catastrophic failures.
The lesson learned in this latest campaign is that it is far better to have analytics in house because the intelligence is likely to be more useful. By insourcing you get not only better quality, you get more actionable results because insiders understand better what is actionable.
I’ll leave you with one more Sun Tzu quote to ponder:
Spies cannot be usefully employed without a certain intuitive sagacity; they cannot be properly managed without benevolence and straight forwardness. Without subtle ingenuity of one cannot make certain the truth of their reports.
Or, netting this out, Tzu says what the Republicans did by outsourcing data analytics was stupid. Learn from their mistake.