A Fashion Retailer and an Insurance Company Meet in a Bar …

I recently spoke with two folks from very different industries with very similar problems. One company was in online and offline high fashion retail and the other was in insurance. Very different industries, very different companies, very similar challenges.

They both told me:

1. “Our company likes to acquire other companies and these acquisitions make a mess of our product and customer databases.”
2. “It is difficult to know when the same product or customer is actually the same product or customer in different databases.”
3. “Our company is a bit antiquated and conservative and we still pull data from mainframes. We are not yet on the cloud, though it may be a good idea.”

The insurance company had a focused IT department and was slowly moving forward. They had already built a functional and partially normalized data lake on premises. Their next step was to hunt down the consolidated data definitions of key performance indicators one by one across their many databases.  A good plan.

The online retailer was suffering with the same problems of merged companies, products and customers but also suffered from inadequate computer resources, out-of-date software and insufficient storage. It was not clear where or when they could begin to fix these problems since any improvements would cause disruption to their current business processes which were already pushed to their performance limits.

While the insurance company had similar challenges to the retail company, they had put together an incremental plan for eventually unifying their databases. The retailer had not. Though both companies were ideal candidates for a cloud-based solution neither had plans for when or which data would be moved to the cloud. 

The agony of acquisitions

Both the retail and the insurance company were among the dominant players in their respective industries. Part of this success was due to their dramatic growth through the acquisition of competitive and synergistic companies. They had both acquired more than a dozen different companies over the previous decade.

This acquisition strategy was an important reason for the companies’ success but it played havoc with the database. A single version of the truth became nearly impossible to maintain.

For the retailer they now had products from different acquired retailers that needed to be accommodated in the existing online store. These products were coming in from many manufacturers from around the world and from multiple brands. The same product might be listed under multiple UPCs from the manufacturer or the manufacturer may have retired a given UPC and then assigned it to a new product. This could result in the same UPC in the retailer’s database pointing to both a pair of pants and an umbrella.

IT activation energy

The challenge for both of these companies was in overcoming the momentum or friction of their current database solutions and to move to a longer term and more elastic solution.

Do you remember much of what you learned in your high school physics and chemistry classes? Probably not – I had to look this up as well but there is something called an ‘activation energy’.  It is the amount of energy or effort required to move a current chemical state into another more desirable one. An example would be the effort of striking a match on the side of a matchbox in order to create enough heat to light the match.

Activation energy is a good metaphor for the situation these two companies found themselves in. They wanted to reach a final more desirable state but they just couldn’t ‘get over the hump’ and provide the necessary effort to move to the cloud.

This was the state that these two companies found themselves in. Knowing that a cloud solution would provide new opportunities to increase revenue and cut costs but they just didn’t have the time and resources to commit to getting over the hump. Once on the cloud though, the activation energy to make future changes would be very low – possibly just the need to spend more money.

The ecstasy of elasticity

We hear about ‘elasticity’ as a benefit of cloud computing all the time but its value was really made clear to me when talking to this online retail product manager. Elasticity solves this problem of IT activation energy.

One of the benefits of elasticity comes from the ability to easily increase or decrease compute and storage to match need at any given time. With elasticity, a system can handle peak need to optimize revenue as well as scale down to lesser priced resources when the need is less.

For this online retailer, the need was even more basic. They had old hardware and software and getting their internal IT group to upgrade was a big deal and projects kept getting delayed. As a result, performance was slow and the users of the system had to wait and watch the ‘blinking boxes’ march slowly across their screen as their data loaded and was processed.

This meant that the people in the company were often prevented from getting their work done as quickly as they needed to. Posting a new product to the online store should have happened within a business day but, more often than not, it took more than a week.

If their systems were on the cloud they could have easily increased the performance. And they would have gladly paid for that increased performance. They needed elasticity.

Four years from now

The challenge of rationalizing products and customers that are inherited through acquisitions is a surprisingly common story for many companies. Maybe this is because the most successful companies are also the most likely to have grown through acquisition.

These problems are particularly painful when you have a data infrastructure that is not elastic and has a high activation energy required for change.

The elasticity of the cloud would dramatically lower this activation energy for IT and allow for many new and powerful business opportunities to be pursued. Expect that early adopters of the cloud will be able to do the following in 2021, four years from now:

1. Predict and then reallocate inventory between physical and online stores
2. Recognize the same customer across business units that were created through corporate acquisitions
3. Upgrade hardware and software in a continuous, automated and non-disruptive process
4. Recognize the same product across business users and reuse sales descriptions and other product metadata between channels
5. Drive business to the physical store through online interventions and be able to measure the impact (Google has already started to do this by tying customer purchase information from credit card data)
6. Create mobile apps that recognize the consumer’s presence at a physical store location and provide timely marketing and sales messaging

One of the great benefits of the cloud is its unleashing of the “infrastructureless culture” and reducing or removing the IT activation energy required for change. Companies that focus on building a new core competency in cloud computing platforms will recognize great and recurring benefits in the near future.

More Information

https://www.eckerson.com/articles/deep-learning-ai-and-privacy

https://www.eckerson.com/articles/enterprise-grade-data-science

http://www.eckerson.com/articles/should-i-outsource-all-data-science-functions-to-just-one-vendor

https://www.eckerson.com/articles/cloud-data-warehousing-producing-the-infrastructureless-culture

http://www.eckerson.com/articles/options-for-building-a-predictive-analytics-program