This is an adaptation of many stories that we have seen in real life. Names and numbers are fictional and any resemblance is purely coincidental.
It’s a rather warm day, unusually warm for the time of the year. Andy, CEO of MoonSun Enterprises, a $2B machine tool conglomerate, peers at the annual results on his laptop. Every metric is trending downward. The equipment sales are down for 7 quarters in a row. Not only is he going to miss his bonus again, but he has to face the board next week. He has no clue how he is going to defend it. He desperately needs a plan.
Andy messages his CFO Mike to join him in his chamber. He just wants to find a bright spot in the otherwise red spreadsheet.
“Mike, you saw the numbers,” Andy glares at Mike’s face and then peers to infinity through the window behind Mike.
“Yes, I have done those numbers,” Mike is confused with what kind of question is this!
“Is there something to feel good about?”
“I am sorry, I think not,” Mike replies, and then pauses for a second. “Actually, the interesting thing is that services and part sales have held up quite well, even showing growth.”
“Who sells them?”Andy queries.
“Not sure, but they are all part of John’s organization.”
A couple of minutes later, John is sitting in front of Andy. He knows the numbers are not good. Mike is also in the room so he feels assured Andy is not going to fire him, not at the moment.
“Who sells parts and services?”Andy asks John.
“Part sales are done by our sales engineer. When sales folks inquire about new equipment needs, they also ask if they need parts?” John pauses and resumes, “We have made it mandatory to ask for it.” He had a glimmer of a smile assuming Andy was going to be happy with this smart move. “And as far as services go, when support or service requests are raised, technicians respond to that.”
“You mean there is no one dedicated to selling parts and services”, Andy takes the conversation to a different direction.
“Why do you need a dedicated team? Most customers will ask for parts when they need it and for service as needed. Why do we have to dedicate sales folks to this? Also, it is not possible to call 50000+ customers and ask them if they need Brass ring 5430 or spindle AT540.” John is a little agitated with the naivety of the question.
John is a pretty smart guy. It’s not everyone’s cup of tea to be the Chief Sales Officer of MoonSun enterprise. He defended himself in front of Andy but the question kept ringing in his head. Is there a way to predict if someone is going to need an AT540 spindle? His daughter is doing her graduation in data science and analytics and he has heard about wonders that data analytics can do. He schedules a meeting with Vijay, CIO who heads the whole IT. Vijay and he go along very well. Both are cat lovers and marathon runners.
On the scheduled day, John and Vijay meet in a conference room accompanied by a couple of their team members. John did extensive discussions with his team as well as a couple of phone calls with Vijay and his team so that they can utilize the meeting in a more productive way. He also updated Andy on this effort. Andy liked the idea. He wants the parts and service charts on his spreadsheet to go up with a higher slope and he knows it is possible based on some of his recent discussions with his peers. Most of them vouched for the aftermarket. If people have their equipment, they ought to buy parts, consumables, and services. And if they are not buying it from an OEM, they are for sure buying from somewhere else.
“Vijay, we need a system that can predict when people would need parts or services for the equipment that we have sold. All the equipment data and transaction history should be there in our ERP.”
“John, we have data in our ERPs but we don’t have one ERP. We have done a couple of acquisitions and with that comes another set of ERP, CRM, and a couple of other systems. We want to integrate everything and do everything from scratch but we never get the approval for the budget. We also replaced our old ERP with a new one last year. All our old data is in old ERP. We want to move it to a new system, but there is no time. Management thinks we have blown the budget way beyond what was needed even for our recent ERP implementation,” Russell, second to Vijay, was ready to vent his frustration and did it at the first opportunity.
“Can’t we just bring all the data for analytics in one place? How hard could that be? There should be tools and vendors who can do that!”John still is not sure if he is grasping the enormity of the problem.
“We can do that in-house!” Russell never liked the idea of buying tools. He feels everything should be done in-house. And for a reason, he has written all those initial MIS systems 3 decades back. He even wanted to write the ERP of the company in-house. They spent $5 million and 3 years trying to write it when someone in management realized the futility of the effort and forced the IT team to buy a standard ERP. Russell still created a lot of roadblocks while implementation and that is where Vijay was brought in to head the IT.
“Vijay, I am not sure about all this data. We just want to know if there is a way to find how among our 50000+ customers, can we predict who will buy parts or need a service. We cannot call everyone every week and that too when we have no clue to whom to call. We have been mostly doing it reactively when someone calls.” John makes his ask crisp and indicates that he wants to remain focused on his problem.
“Vijay, can we find a tool in the market that can take all our data and generate these insights,” John looks at Vijay. He does not want to engage with Russell. He and his team don’t have a good experience to vouch for. Every other data analytics that they have asked for takes ages. Russell even wanted to write the CRM in-house. John shot that idea and they went ahead with a standard tool.
“Why do we need a tool? We can do it inside. It is all about putting the data in a spreadsheet and running macros.” Russell jumps even before Vijay can speak.
“How much time will it take?” John looks at Russell without blinking.
“Should not take more than a week,” Russell very confidently replies. Vijay does not like doing commitment without understanding but he has never figured out how to control Russell.
A couple of months pass by. Russell keeps promising next week.
One day John loses his patience and schedules a meeting with Vijay and Russell. Vijay has meanwhile taken John into confidence by telling him that he will move Russell out of all other activities so that he can focus on it completely but he will not own the commitment on Russell’s behalf. John was fine with that as long as Russell worked on it full-time.
“Russell, it’s been 15 weeks now, and we have not seen a single result yet!” John comes straight to the point.
“Oh!, you know, it is all an Excel problem. Microsoft has not done a good job! It does not know how to load so much data. Also, the data is very bad. I tried to clean some but it was not possible to do for such a large data set. And this is for just our current ERP.” Russell explains his position.
“Russell, just looking at the data of the current ERP is not that valuable. It is hardly 3 years old. These machines will be new and will not need as many parts or services as the older ones. We should really look into old data,” He looks at Vijay, “Can’t we buy something off the shelf? We would have been up and running by now. I am probably losing a million-dollar business every week!”
Vijay though is a sensible guy but buying something from outside usually is a hard thing even to push in his team. And it’s not that all bought things have been successful. He has his own share of failures. He was not sure initially if it is a simple problem that could be handled inside or if it needed specialized partners. But now he has started realizing the complexity of the problem. It is not trivial. Data from almost 6 ERPs and 3 CRMs. They have one FSM system as well. The data is sitting in the mainframe and he knows many spreadsheets are flying across the organization which contains gold mines. Some of the recent discussions with John and his team are clearly giving him the idea that putting all the installed base information at one place can create wonders. It is not just about an MDM but something that is smarter than MDM and runs AI and analytics on top of this single source of truth and predicts the parts and services needed.
“Russell, is there a way we can fast-forward it?” He wanted to help John sooner than later. And on the back of his mind, he knew any purchasing decision would take another couple of months. John meanwhile searches for “Installed Base selling” and finds a company that is claiming that it can find selling opportunities over the Installed Base. John shows it to both Vijay and Russell. “Should we talk to them at least?”
To Vijay, the offering by the vendor looks interesting. He browses it a little more and finds that they even do the single source of truth for all the installed bases and can pull the data from mainframes as well. They already have AI algorithms.
“Must be very costly,” Vijay murmurs.
“Probably we should check with them before we go further in our rat hole.” John is desperate for results.
“I think we should just hire a data scientist. I know everything that needs to be done. Data scientists also clean the data.” Russell intervenes as he sees the discussion moving in a different direction.
“But that will be a costly hire.”Vijay is not sure if he is ready to foot the bill yet.
“Russell, why don’t you check with this vendor as well first? They might just help us out.” John brings the conversation back to the original trail.
“I will do it,” Russell picks up the offer.
Next few weeks he informs both John and Vijay that he tried to reach the vendor but no one responded. In reality, he just gave a cold shoulder to the vendor. He wants to build it internally, at any cost. Meanwhile, he also convinces John and Vijay to get a data scientist.
The next 3 years go in Giffy.
Data scientist comes and complains about the data quality first. When Russell asks him to clean, she tells him that as a data scientist, they clean a limited set of data to work on the model. Data scientists are not data cleaners. So they go ahead and buy a data lake where they can pull all the data. 3 people are hired to manage the data lake and who can write integrations so that the data can be pulled. In no time, the team grew to 5 people. Meanwhile, the data science team grew to 3. Still, no real work starts. The data is somehow pulled into the data lake but it is like all the data sitting next to each other. No better than when the data is sitting in the original system. They still have no clue how the incremental data will be fetched after the initial load. The team has gone up from 5 to 7 to handle all this. Someone realizes that they have still not figured out how to clean. In no time, a team of 10 data analysts appears. Their work is to go after each record and clean it by looking at the internet. Still, it is a long way when they can even think of deduplication. The team at best is able to clean some addresses. With no standard part catalogs and when data is going decades back, they have no clue how to clean part names.
Someone suggests third-party tools.. All kinds of tools are purchased from data preparation to deduplication and dashboard tools. With no notion of a data pipeline, the data is literally manually loaded from one tool to another. With each tool, came the need to hire many of the tool experts. Russell has a team of 30 people by the end of the first year reporting to him. They even purchased a cloud provider to host various things. There was a suggestion to buy one more cloud provider service as they were giving huge discounts in the first year of operation. After the first year, the cost silently adds to bills.
Mike, the CFO, with activity-based costing, puts a tag of $3 million for the first year and a total attributable investment of approximately $12 million to date. Let’s not talk about opportunity cost!
People used to build cars in their garages a century back, but not anymore. There are some aficionados who still do it. Car manufacturing has become a very specialized skill and needs corporations of huge scale to build cars in an institutional and automated way. The same goes for software. It was easy to do them in-house and at an individual level two decades back. These are no longer grandma recipes. The systems are more complex and need specialized teams to execute them in an institutional way. Beyond building software, there are issues around data governance, reliability, and performance, each of the topics needs its own specialist.
John and his team are still looking for results. Even after 3 years, Russell keeps the promise of delivering next month. They do get some results but with all the complex structure in place, it comes with a lag and remains not so useful. The results are at times delivered in spreadsheets or as dashboards. Not being part of their daily workflows, the results are not so actionable. With time lag they become stale before even the team can start looking into that. There are no easy-to-use interfaces to interact with data. In the world of web apps and mobile apps, this is the least expected.
John just had a heated discussion with Russell in his cabin. Vijay remains silent throughout the discussion. Andy was fired last year and he believes he will be next. The new CEO is evaluating all the investments and ROI minutely. Vijay from his experience knows that build vs buy are not straightforward decisions. There are situations where building in-house is more important. Going wrong on buy is usually not costly as one can just cut the thread, however, organizations can go monumentally wrong on the build side and the mess is never easy to clean. He still remembers the ERP mess when he joined the MoonSun. Russell could survive because of an old-timer and has some deep connections in the organization.
John types on the search engine “Installed base selling”. The first search is about a company that provides an Installed base platform. He now needs a reliable partner to execute the idea. He remembers one of his friends, who heads the service organization of a large packaging equipment company, has told him about this company. He fills out the contact form
John Travola
john.travola@moonsun.com
In a couple of follow-up calls, he gets an estimate of the solution and he realizes it is on the lower end of 6 figures. They even promise the full running implementation in 10-12 weeks. He holds his head in despair and blames himself for letting Russell talk to the vendor.
He has to move fast.
Read about Your prospects’ dilemma – How do they make a case when there’s no viable competitor?
Gain valuable insights through the life stories of Industry Experts, and subscribe to the Aftermarket Champions podcast.
