These words actual mean something to the industry and empower teams to coalesce around the technologies and legal frameworks necessary to participate.
Adtech, edtech, fintech, etc.
each have their own basic data dictionary.
If you don’t speak the language you could be having the wrong conversation.
Throughout my career I have joined doomed data monetization conversations near the point of completion.
A data provider had a desirable suite of audience intelligence but the team had not yet figured out how the segmentation would be integrated into the application.
Months of business development vetting quality and scale wasted because there was no way to connect the dots.
The gaps could be technical, legal, financial or all of the above.
It only takes one chasm to kill a deal.
Luckily, these issues can almost always be identified early with a simple call flow diagram labeled with defined data elements.
You do not need to be an engineer to put one of these together.
The goal is to start a pragmatic conversation not document the architecture.
Call Flow for Weather Specific Ad CreativeWhen we were asked to design an ad that dynamically updates based on current weather, our team needed to integrate with a weather data service.
The IP address was the signal needed to resolve a zip code that we passed to the weather service.
They responded with a segment, a two digit weather code.
The creative team stored these codes in the weather ads and used them to dynamically update the creative of the ad to reflect the current weather.
It was simple and it worked.
There was no need for a unique consumer identifier in this example.
As the number of hops between data points increases, so does the value of these diagrams.
Most data applications, like bringing mobile carrier segments to advertisers, require a user identifier and are not so simple.
A billion dollar deal will not pay off if the dots only connect under rare circumstances.
Bringing product and engineering into the conversation early is critical.
The key to success here is to use a commonly accepted data language.
Your industry already has one — learn it.
If you think you have invented a new type of data or a unique use of data, take the time to work with your team and partners to clearly define what it is and what it isn’t.
Create a naming convention, stick to it, and market it.
Against popular belief, there is no easy way to make money with data.
Value extraction is only possible when an organization takes the time to define its data assets and learns how to talk about them internally and eventually with customers.
Know the ProblemData conversations have a tendency to begin and end with arguments about the scale.
Don’t let scale immediately dissuade you from pursuing a data powered product.
The market will always value scale, but customers will help determine the right balance between scale and precision.
The important thing is to find a problem worth solving.
We experienced this at Jumptap and learned key lessons on data monetization along the way; know the problem, know your customers, and then identify the right scale.
In 2011, there was no consumer purchase data available to advertisers targeting mobile devices.
To reach an audience segment an advertiser had to trust a publisher’s contextual meta-data and hope for the best.
The Jumptap leadership team recognized this market need and we began designing a way to bring consumer behavioral data into mobile advertising.
Our first attempt to reach targeted audiences on mobile devices focused on getting to scale fast.
We licensed consumer purchase data from DataLogix (now owned by Oracle).
They proved to us that certain postal codes displayed above average purchase behavior for certain product types.
We could already target predefined groups of zip codes, so we created specialized location targets based on the DataLogix purchase data and started selling these location based audiences to advertisers.
After initial success, customers began to question the precision of the product.
They expected a one-to-one match between the purchase behavior and the devices.
Our solution did not offer that.
Our second attempt valued precision over scale.
Data providers could send us audience segments for any user for which we had an email address.
We reached out to our publisher partners and asked them to collect emails on their users and to send them to us after putting an obfuscating hash on the address.
We paid publishers for millions of hashed emails linked to valid mobile device ids and proceeded to onboard data from our data partners.
The more precise product was reasonably successful, but this time the pushback was on the lack of scale.
We were never able to get enough email addresses to serve the needs of our customers.
The Complex Map of Consumer Purchase DataFinally, we found a compromise that balanced scale and precision.
We figured out how to link mobile devices to specific households with accuracy.
The targeted advertising was then delivered to all of the devices we could reach that were linked to the household.
This scaled well.
Our customers recognized that influencing purchasing behavior was highly effective when the entire household was getting a similar message.
We immediately tapped into ad budgets that were constrained by the lack of targeting available in mobile advertising at the time.
Then we took the same technology stack and used it to power some of the earliest cross-device targeting and measurement solutions in market.
Our ability to deliver a product at scale while maintaining a market validated level of precision was a key contributor to Jumptap’s ultimate sale to Millennial Media for over $200 Million.
We knew the problem we were trying to solve, we listened to our customers, and we managed the trade offs between scale and precision.
Have the ConversationEvery organization creates data as a natural course of business.
At some point they arrive at the conclusion that the data might have some monetary value.
Over the years, I have developed a framework to help drive the conversation around data monetization to help teams figure out if it’s worth pursuing and how to align the team behind the decision.
The first step is covered in the the beginning of this article.
An organization has to define the dataset being considered, confirm that the organization has the legal right to use the data, and estimate the current scale.
In my experience, 9 out of 10 companies have to go back and rewrite some data use or privacy language before executing on a monetization plan.
It takes time and it can be controversial.
Do this first or pay the price later.
The next step is to explore the monetization opportunity by assessing the data’s potential to be refined and externalized.
The matrix below is a good tool to drive the conversation.
All data starts in the bottom left — raw, internal and not creating any value.
The matrix forces the team to think about how a dataset could move into a different box.
Up, as it is refined, and to the right as it is shared or sold.
Do we have unique access to a data type?Does the data resemble raw material or a finished product?Can we make any of our data available to other companies?Do we want to invest in refining our data?Does a market exist for our data already?Is there anything differentiating about what we do with our data?Answering these initial questions and placing data assets into current and aspirational boxes helps align the team behind a broader data strategy.
To help illustrate the options a little better, I put company names into their primary boxes below based on their current monetization conversation status.
A company can have products in multiple boxes.
Amazon had the conversation and concluded that their strategy was to create a Fort Knox around their data asset while innovating with data refinements and internal data services.
They invested heavily in data democratization services to empower their employees.
They use customer intelligence to grow their business and have been wildly successful.
They are now the model of a data driven enterprise.
Seeking to be the next Amazon is not a bad decision.
However, if leadership decides to focus internally they need to communicate and defend that decision periodically.
Innovative and entrepreneurial employees will identify commercial opportunities worth considering.
Don’t be afraid to have the conversation.
Offering Data as a Service (DaaS) is a massive opportunity today.
Companies that have a competitive advantage acquiring, delivering, or refining data, like Datalogix (acquired by Oracle), and The Weather Company (acquired by IBM) have the potential to reach billion dollar valuations.
Going down this path is a major strategic decision that requires significant investment.
If this is the output of the monetization conversation, getting aligned and building out a detailed plan is the required next step.
Try to stay in one box, and put a huge emphasis on differentiation.
PlaceIQ, the other DaaS company in the matrix, was one of the early visionaries in the location data ecosystem but they had challenges establishing scale and differentiation.
Over the years they have developed products in practically every box before finding success with their current platform providing highly refined location insights for customers.
The final column, the data sellers, are dealing with bad press these days thanks to Facebook (not a seller) and other actors that have been less than transparent with their consumer data practices and policies.
Nevertheless, selling data is a legitimate business that can be done right.
Long before the Internet, the original data vendors started selling refined datasets like household demographics, purchase histories, or other high value lists used for various purposes.
These grandfathers of data still sell shrink wrapped software.
It is not where the data business is heading, but for some datasets it is still a valid business model to be considered.
A new crop of data sellers have identified that there are digital markets for all sorts of raw and refined data types.
In the US, except for financial and health verticals, the exchange of data is generally unregulated.
This might change in the future, but today companies like Oracle, Lotame and Dawex provide generic ways to monetize defined data assets.
Niche buyers, like location intelligence provider Cuebiq, purchase raw location signals from companies like GasBuddy.
As the matrix suggests, selling raw data into the data ecosystem is a low position on the value chain.
Nevertheless, starting here is a perfectly valid way to test the value of data if your company is considering the move to a DaaS business or becoming a data consultant.
Overwhelmingly, I see experts argue against selling data but I think companies need to take a pragmatic approach.
Have the monetization conversation.
If your company has full rights to the data asset, limited exposure to risk, and no short-term intention to build out a data business, then participating in the data ecosystem should be considered.
Use the diversified revenue stream to build up enterprise value in other ways or move up the data value chain over time.
The monetization conversation is neither a one-and-done exercise nor is it a treadmill that you want to be on forever.
Every company needs to figure out the right cadence.
If your company goes down the Amazon path, you should expect to reassess this decision periodically but not every month.
If an opportunity surfaces, test the market.
The shelf life of data today is so short that a test engagement can provide a lot of learning with limited exposure.
Most importantly, as the team assesses monetization opportunities, take the time to document opportunities, aversions, and results.
Having a paper-trail of the conversation can help avoid deliberating the same question repetitively.
It can also help an innovative employee identify an opportunity that leadership has not yet considered.
Demand IntegrityMonetizing data at scale is not easy.
For every Datalogix that reaches a billion dollar valuation there are a hundred companies like Rocketfuel that find a unique application for data but make critical missteps.
To be successful you need a team that has tremendous integrity and the ability to communicate the complex.
I pick these two traits because with today’s technology you could hypothetically do anything with data.
This fact makes it too easy to exaggerate capabilities and underestimate complexity.
The team needs to communicate the value clearly and then deliver.
The results (expressed as data) will ultimately uncover the truth.
Do not risk faking it.
The challenge is that hyper-competition will constantly push teams to try to do more and more with their data.
Investing in systems for security, privacy and policy training are important.
None of these make up for leadership that exudes integrity around data usage.
Regulations like the European GDPR are the result of technology hubris and years of opaque data collection.
It would be a tragedy if we let this slow our progress.
The solution is simple.
Teams need to embrace the new norm of data transparency, consent, and control.
It is a fair balance.
As companies profit from their use of data they also need to provide consumers with greater access and control over that data.
If an organization expects to maintain a data drive advantage, management has to make this new paradigm part of the company culture.
Examples of innovative data uses are all around us.
From the Google search that brought you to this page, to the smooth ride in your car because your local government is crowdsourcing pothole locations, data is used to make our lives better.
The opportunities are endless, but it is easy to get overwhelmed by the risk and complexity associated with building these products.
Objections are easy to find when the effort feels gigantic or the data asset unobtainable.
My experience has proved to me that there is tremendous value trapped in enterprise data.
The key is to break down the big problem into its minute data points using plain language, have the monetization conversation to get the team aligned, then communicate with brutal honesty as the team delivers.
Following this framework and identifying where your company has a data opportunity could be the difference between growth and stagnation.
That is the final lesson.
.. More details