A WeDo Conference Talk: Consulting & Analytics: Improving your Business Today, Enhancing it Tomorrow

At Miami Beach, I recently attended an excellent risk management conference put on by WeDo Technologies and Mobileum.

One of the most informative presentations at this “We Meet Americas 2019” conference was given by the consulting team, led by Thomas Steagall, WeDo’s Vice President of North America.

The session covered the following:

• The framework for WeDo’s risk management consulting;
• A walk-through of a few consulting engagements in areas like subscription fraud and collections;
• Examples of how analytics and machine learning can supplement process optimization; and,
• An overview of WAVE, a new, post-consulting program where WeDo software tools and data scientists are brought in on a month-to-month basis to help a carrier achieve certain KPI targets.

I lightly-edited this talk, but it’s otherwise presented the way it was delivered to the audience of about 130 people, including delegates from 30 operators.

Consulting & Analytics:
Improving your Business Today,
Enhancing it Tomorrow

September 26, 2019
Fontainebleau Hotel, Miami Beach

Thomas Steagall: Friends, there’s no book you can read that will tell you exactly how to manage the business risks to telecom operators.  It’s up to Risk Managers to take their decades of knowledge and experience in the business and put it to work.

But even then, you run into unexpected troubles.  You lose customers.  You lose revenue.  And you sometimes overpay.  These problems have always been there.

But now, your job has gotten much tougher because telecom has become a highly dynamic business that’s bringing fresh challenges:

The Risk Challenges in Telecom Today

• New Mobile Technologies are Being Deployed — Ericsson, here at the conference, spoke about how it’s deploying 5G, and even 4G LTE is still being digested in some markets of North America and Europe.

• Massive Numbers of IoT Devices are Being Connected — We not only have subscribers talking on phones, mobile devices, we now have machines talking to machines — and machines talking to humans.  So everything that surrounds us will be connected to that very same network, constantly exchanging information.  From a risk perspective, this injects a gazillion new variables into the game — and these are things we are not used to.
• Monitoring Risks for Different Industries — Particularly as IoT emerges, telcos are being asked to monitor risks for companies in different industries — and that is a totally different ball game.
• Customer Expectations are Raised in the Digital Age — Our market also requires us to digitalize the front end towards our customers.  Customers expect us to enable countless tasks from their phones: from paying bills... to monitoring their home security — and pretty much everything else.  The customer doesn’t want to get on the phone to talk to another person about their problem.  They want their smartphone text, email and chat interactions to solve their problems in real time.
• Fraudsters Aim to Exploit Operational Complexity — As if you didn’t have enough to worry about, the familiar bad actors are out there trying to find loopholes in all of our complex processes to steal money from us and our customer.

So, these challenges are quite immense.  But our consulting team brings a lot of expertise — operational expertise — to many different dimensions of business assurance: both the non-intentional risks of revenue assurance or business assurance, and the intentional risks of fraud, which crossover into security, and also includes the risks of the sales chain towards commissions and dealers.

Still, it’s rare for us to hear about telecom business situations we haven’t seen someplace else before.  When we do, we learn something new, and it gives us bright ideas on how to tackle those things going forward.

WeDo's Consulting Framework

Now our consulting team very much works from a framework: an initial assessment and overview of your operations.  That gives us the blueprint of:

1. Where are you today?  What have been your team accomplishments with your processes and organization?
2. What are the gaps and areas of opportunity that are out there for you to explore.
3. What are the potential savings or risks to your bottom line?
4. How do we go about closing the gaps, recovering money that was left on the table or left on the floor, and getting better — getting ahead of the problem that you have?

In a nutshell, that is what our team does.  And our brilliant data scientist, Carla, is now going to talk about how she collaborates with us in deploying AI in use cases that touch operations.

Optimizing the Collections Process

Carla Cardoso: Thanks, Thomas.  I’m going to highlight a collections case which shows how clients often need a mix of services to solve their problem: both consulting to improve processes, tools to improve the operation, and finally analytics and machine learning tools to deliver on-going insights, automation and constant adaptation.

At WeDo, we address certain collection management problems through solutions such as our RAID Collections tool.  This will help automate operations and lead to cost reductions.  One of the handy things it does is provide business rules that are automatically run every day to make things more efficient.

But what if there are deeper issues than simply saving costs?  What if there are problems in the very policies that guide the collections function?

Obviously we’d like to anticipate bad debt, prevent it, and make our collection efforts more successful.  If we detect a risky prospective customer, we want to block him from becoming a customer.  And if he does enter, we want to lessen the damage.

But there’s another key factor to consider.  Many customers are good payers who, on occasion, run into financial problems.  So if we treat them the same as the bad payers, we will create a bad customer experience.

So collections is not just about costs: it’s also about treating customers right.  And this is where machine learning can prove very useful.

And while automated systems can help, there’s also a need to fine tune policies.  Strange as it may seem, an operator’s collection policies often incentivize customers not to pay.  Customers end up not paying for two months knowing that nothing will happen except getting a cordial call asking them to pay.

So reviewing collections and billing policies often brings problems to light, and these are not problems a software product will solve.

As Thomas said, it’s important to evaluate everything, having a holistic overview of an operator’s problem so we can recommend steps to optimize operations.  So this is the difference between going in with a product-oriented approach aimed at selling you a tool versus having a bigger picture and selling you services to optimize operations.

Still, analytics and machine learning can deliver great value because they allow you to examine details across the full collections cycle from bill to debt recovery.

For instance, analytics can calculate the risk of non-payment, so we’ll know when we are likely to receive payment from a particular customer: in 5 days, 10 days, or 20 days.  Then, when a customer goes into debt, we can predict: What is the best channel of communication with him?  What is the best time of day to contact?  How frequently should I contact?

More proactively, we can help prevent debt by determining the best bill cycle for each customer.  If we can adjust the date of payment to please individual customers, you suddenly find more customers are paying on time.

For me, machine learning and analytics are like tools in the kitchen.  You use them where it makes sense.  Bring me a recipe and I will decide what tools in the kitchen I will use to make the cooking job easier.  It all depends on the ingredients — the data or the particular business problem at hand.

Thomas Steagall: Fantastic.  So, here is a consulting case with an operator in the Caribbean.

Fraud & RA Problems at a Caribbean Operator

The Problem: A Caribbean operator had a small RA/fraud team, and they were getting hit with pretty severe losses in different parts of their ecosystem — and these attacks were a big surprise to them.

Financial Assessment: At the request of the CFO, we came in and did what we call a financial risk assessment.  That gave us a picture of their entire organization, even outside of RA and fraud — how their business was operating, how IT was collaborating with network and engineering, etc.

And based on that assessment, we said, “There’s an opportunity here to recover X million dollars, and we believe that we can help you materialize some of those savings.”

Six Month Investigation & Savings Recovery: Right after we presented that report, the CFO gave us authority to pursue those findings, to investigate them, prove that they actually exist, and understand why they happened.

Why did you lose money?  Why did you have such a severe fraud case happening in your customer acquisition process?  More importantly, how do you fix that and drive recovery?

The results were actually surprising: we discovered 20% more money than we originally predicted.

Money Recovery Engagement: During the project, our consultants performed their analyses and created the design for automation of their controls.  So the CFO was very happy because it became almost a self-funding process.

From the money we recovered for him in six months, he could pay for the tool and pay for our consulting.  Plus since everything was automated — and it was deployed smoothly using an agile process — so it didn’t take a lot to convince the CFO to continue our work because we were generating big savings.

A key point to emphasize here: Every time we find leakages in fraud, it’s not always a systemic failure.  Most of the time it is a process problem, or people problem, or governance problem.  It is mostly a lack of control — and not always an IT issue.

Now here’s another subscription fraud success story where Carla was deeply involved.  Carla, would you please talk about that too?

In-Store Mobile Handset Subscription Fraud Control

Carla Cardoso: Yes, in this case, a CSP was having critical losses related to Device Subscription Fraud.  The company had developed its own metrics to score subscription risks, but that approach wasn’t working as the operator was losing money in in-store handset sales.

What we did was construct analytic and machine learning models to score the risk of each individual activation.  So if the prospect was trying to activate four accounts and buy three iPhones, then the system might calculate the fraud risk at 90%.

And the system would recommend alternative offers, such as the customer signing up for a less expensive iPhone contract.  So in this way, the agent didn’t have to totally reject the sale but adjusts the offer to align with the risks.

And as more information was gathered from the prospect during the sales process, the risk model was adjusted accordingly.  It would also give hints to the salesperson to guide the sale to a win/win solution where the sale is made within reasonable risk for the CSP.

WeDo’s WAVE: Professional Services & Tool Deployment

Thomas Steagall: Great, Carla.  So as this brief presentation has shown, we have pretty deep know-how on the WeDo consulting team in risk management, and have great experience solving some of the process issues as we’ve discussed.

Now most of you are clients already, so you also know our software and technology.  This year we launched WAVE, which is our approach to business managed services, where Consulting, Products and Analytics are connected.  And here to talk about that is Bernardo.

Bernardo Lucas: Thanks, Thomas.  WAVE includes some professional managed services combined with deploying our tool, which we tailor to meet your specific operation.

Now we are not talking here about a pure BPO, Business Process Outsourcing of everything.  No, this is a selective area of the business where we bring value and work against results.

Once you get to the stage where the WeDo consulting project is finishing up, there may be a specific aspect of the business where you want some on-going help.

For instance, in the subscription fraud area, the customer asks us, “Can you help me manage this on an ongoing basis?” And that’s when the WAVE team brings in our technology.

What happens next is WeDo integrates all of the data with your point of sale system — and with a data scientist on our team actively building the solution.  And then we tie this against a specific results-achieved KPI which is our target.

The WAVE tool is deployed and billed on a monthly service basis.  When we achieve the KPI goal, WeDo is paid a success fee for reaching that target.  Then, as time goes on, the data scientist continues to evolve the models to keep the success going.

So that’s what WAVE is all about.

Thomas Steagall: Thanks, Bernardo.  So I think we’ve given you a full picture of WeDo’s professional services program.  First, our professional consulting services helps you assess your business and optimize your risk processes.  And then, if you need on-going services after our consulting is complete, that’s where WAVE’s tools and data scientists come into play.