Why Should You Negotiate for Data?

Source: Rawpixels_Pexel

Source: Rawpixels_Pexel

Don’t bargain yourself down before you get to the table.
— Carol Frohlinger

Next time you negotiate with a strategic customer, ask them to include more data. To understand the value of data, imagine you are a director of a radiology company with millions of digital images. A company asks to buy 100,000 images plus your diagnosis of each image.  This company will use these images to train Artificial Intelligence to analyse images automatically.  How much will you sell them for?

This is a tricky question. Some people would ask themselves: how much would it cost to transfer the data. Just hundreds of dollars, say $200, but this seems too cheap. Even allowing for removing personal details and getting some legal advice, maybe $2,000. Still, this seems too cheap.

But consider how much it costs to create the data. If you keep things simple and say, each diagnosis took a radiologist one hour and value a radiologist’s time at $100 an hour. So, for 100,000 records you should charge $1 Million. 

Now you have three possible prices: $200, $2,000 and $1,000,000. How much should you sell the data for?

Focus on Value not Price
— Stephen Kozicki

The answer is to change the question from price to value.  Stop looking inward and start looking outward. Negotiating comprises two processes: Creating Value and Capturing Value.  So far, we have considered how to capture value based on only what we know: what the data cost.

Creating Value for Them

What Value will the data create for the Artificial Intelligence company? In practice, how do you find this? An experienced negotiator, Leo Hindery gives some simple advice: do more homework than the other guy.

  1. Research the Artificial Intelligence company. Are they a monster company like IBM trying to tweak their existing Watson software? Or a tiny startup specialising in radiology?  What’s their revenue and profit? Or what are typical profits for an organisation like theirs?  What are their priorities?

  2. List your assumptions. For example, I assume an IBM will have piles of cash available, or I assume a tiny startup will have little cash.  Both assumptions could be wrong; IBM might have global rules preventing cash deals. In contrast, a tiny start-up may have backing from a giant venture capital company like KKR with plenty of cash.

  3. Plan some open questions to test your assumptions

a.      What will they do with data?

b.     How will the data help them?

c.      What will they be doing in 5 years if this work is successful? (Listen for products and markets, e.g. we will have product X selling in Asia Pacific and America)

d.      How did they decide to approach your company? (Listen for clues you might have something other possible suppliers don’t have, e.g. a more samples of a particular case or some differences in data collected.)

e.      Who else have they approached?

f.        What else could we do to help you be more successful and faster?

Place a higher priority on discovering what a win looks like for the other person.
— Harvey Robbins

Creating Value for Us

You now have some clues on how much value your data may create for them. Now flip your perspective and ask: What value can the Artificial Intelligence company create for your radiology company?

With so much uncertainty about a company and its products, how can your company benefit? A great question. As complexity in the world and complexity in products increases, then in negotiations, it’s certain we will face more uncertainty.  One useful technique is to think about options: in this case, how can our company get a share of possible success. You have an option giving you a possibility of getting nothing if the company fails and getting money if the company succeeds.

Examples of possible value your company could get:

A.     2% annual revenue for all products sold

B.      5 free software licences for 10 years

C.     20 hours of Artificial Intelligence consulting

D.     6 days of Artificial Intelligence training

Imagine if the product becomes a best-selling global product, what would some items be worth? Notice most items could be high value for you and be low cost for them.  Also, as they develop the product: earlier they would value these items lowly; later when the product is proven, they would value these items highly.

So, you may think you could get A or B or C or D. Because now the other side will not see these as high-cost items, we suggest you could easily get A + B + C + D!

Practically, now you can create three packages with different amounts of cash, different royalty rates and different items that create value for your radiology company.

You can create plenty of value from Data: So, ask your Strategic Customer for data more often

Now you can see it’s tough to value data and it’s easy to create more value from data. So, next time you negotiate with a strategic customer, ensure you ask them more often to include more data.