A typical Data Science consulting inquiry
One of the hardest things to sell is Data Science consulting services. That last statement might leave you scratching your head since for the past few years industry has been extolling the virtues of the Data Scientist and what they can do for your bottom line. But here’s how a large number of conversations go:
Company: After our discussions with you we believe we need to incorporate Data Science into our company.
DesignMind: Great! How should we set up the first consulting engagement?
Company: We’ve decided to bring somebody on full time…
In the past two years we have talked ourselves out of about half a dozen engagements due to selling Data Science too well. So we’ve begun to take a new track based off the standard conversation:
Company: After our discussions with you we believe we need to incorporate Data Science into our company.
DesignMind: Great! How should we set up the first consulting engagement?
Company: We have decided to bring a data scientist on full time…
DesignMind: Sounds good. How long will that take you?
Company: …
DesignMind: How will you hire this person, what skills will you look for, and how will you define their goals?
Company: …
What’s the cost of waiting?
The real concern with Data Science engagements is the loss of knowledge after the contract is finished. Nobody thinks about the time delay in gains made by waiting to hire someone, on-board them, and bring them up to speed. So the discussion then turns to the expected return on investment (ROI) on an engagement that lasts until they find a suitable data scientist to bring on board full time.
So, what is the ROI on a Data Science engagement? Our rule of thumb is that an engagement’s ROI should be greater than the investment within six months of the start date. Therefore after six months (or ideally less) you would begin to see a cost saving.
How it often goes
What typically happens is that we are brought into solve a problem. Inherently I find other pieces of the system that can be improved along the way. These pieces that are improved and deployed before the original deliverable is done typically have an ROI greater than the original cost of the engagement.
One of our recent engagements was with a non-profit. They sent out hard mailings every 6-8 weeks. Because of their large mailing list, this was a very costly endeavor. They engaged me to help optimize their mailing rates, and for a two week engagement they had an estimated ROI of almost 20X in just the first year. To give you some numbers, the engagement cost was $12-14K. They originally spent over $1MM per year on mailings, and using our suggestions, they achieved a cost saving of almost 30% ($300K) with less than 1% loss in gain.
And now what?
If you’re wondering how a data scientist could make your organization more efficient and potentially identify cost saving actions, talk to several data scientists about what they recommend. This could include consulting firms or individuals who you may consider hiring for your permanent staff. Ask them for some high level recommendations and time estimates for implementation. To plan your interviews, you may want to check out Hiring a Data Science Team: Hiring Your First Data Scientist. Then, after you’ve talked to several data scientists, you’ll be more prepared to decide how to proceed based on concrete information.
If you take this route, the Data Science team at DesignMind would be glad to meet with you via phone or in person for a one-hour consultation. But whatever you decide, do your homework first.