“The Bear” on Business

Ike Elliott had a good post on churn today, using Vonage’s 1Q08 results as a backdrop.

I started to post a comment, but began rambling.  I’ve been hanging around John Scarano too long I guess.  Anyway, I decided if I was going to poke away at the keyboard, I might as well drive some traffic to my blog instead of Ike’s.  So I will complete comment here.

Ike’s analysis is interesting and useful. I will add one more concept to it.  What is a reasonable expectation for steady-state churn?

A 1% churn rate implies a dollar of revenue will stay on the books 5 years.  Think of all the reasons a customer might churn–go out of business; move; out-grow the service; benefit from repricing; or switch to another service provider.  If, on average, only one of these events every five years would seem to be a a fantastic result for most recurring revenue businesses. Yet many telecom business tout 1% or lower rates and attribute it to their great service.  Eventually, it becomes clear that the churn rate is really higher–this is explained usually by citing external uncontrollable macro events such as a slowdown in the economy. 

What is really happening?  Newer companies should have low churn.  Why?  –the maturity of their customer base is low.  If you have been in business only three years, your average customer has been with you less than 1.5 years.  Your churn better be very low!   Else, you have bigger problems such as bad service or an unsticky product.

Rapidly growing companies have a similar dynamic.  When Cbeyond was in its rapid geographic expansion phase, it’s average customer life was unnaturally low.  Churn, it follows, should be much lower than steady state.  Is this how they explained their low churn numbers in their early years?  I don’t think so.

Tracking and understanding churn is very important in recurring revenue businesses.  I wrote a series of posts on the topic.  For those in the telecom services business, they are worthwhile to read.  Look over at the categories on the left and click on Disconnect Inferno to read more.

Jerry Mills Collins was one of the original members of the Trammps. The following quote is from Wikipedia:

In 2000, group member Jerry Mills Collins was convicted of beating his wife with a handgun on Valentine’s Day when he suspected her of infidelity. Collins was found guilty and sentenced to 12-35 years in prison.

If you don’t believe me, click here to see it.

Given my newfound affinity for Disco Inferno, I pondered what could have caused Jerry to take such a horrible turn in his life. I researched further into it and found this–please click here. It certainly explains why after 1999 he was never the same again. 

This last disconnect post (for a while) will cover a few odds and ends on churn.

1. Establish a clear definition of disconnect: For example, a disconnect is an event where an existing customer no longer needs a particular service thereby causing revenue to be removed from our billable run rate.
2. Include re-rates: It is usually better to retain a customer and lower the price than lose the customer all together. However, the disconnect process must capture the impact of price decreases. It should be categorized differently than a pure disconnect.
3. Disconnect metrics should capture both disconnects and re-rates: Though re-rates are less painful, they still must be included in reports that track the revenue attrition.
4. Include disconnects that are associated with upgrades: In Telecom-land, a customer disconnect is sometimes the result of the customer stepping-up to a higher level of service. For example, the customer replaces an OC-3 with an OC-12. Sometimes these are referred to as good disconnects especially if the amount of revenue is higher after the upgrade. My philosophy is to treat these disconnects just like all others–and treat the new service as a new sale. The net positive impact comes out in how we track net installs (that is, gross installs less gross disconnects.) For those who treat good disconnects differently, I offer a word of caution. Read my prior post again to understand how tempting it is to call murky situations a good disconnect if this term is more socially acceptable than a bad disconnect, thereby creating false security.
5. Verify accuracy of reporting: I’m not sure exactly how our folks do this but I know they do. In general, they run billing reports that compare this month’s recurring revenue to the prior month. Changes are flagged and compared to the disconnects that are tabulated in the managements reports. The goal is that 100% of actual billing changes have been included in management reports–and vice  versa. When there are exceptions, the process is analyzed for what caused the miss and the process is tightened as a result. The goal is two-fold: (1) we don’t want a false sense of security about what is really happening with disconnects, and (2) we want to make sure revenue that should be billed is in fact billed.
6. Set reasonable goals: Generally speaking, a disconnect rate of 1% or less is excellent. (1% means that 99% percent of last month’s revenue was retained during the current month.) 1.2-1.4% is usually acceptable. 1.4 to 1.7% is manageable but not comforting.  Above 1.7% is generally a sign of bigger problems.

With this completed, I hope to never have to write about disconnects again. More importantly, I hope I side-step the next Disconnect Inferno.

A while back, I wrote posts pertaining to how we expect operations to accurately predict when sold (but not yet installed) orders (which we call pipeline) will get installed. Prior to that, I covered how our account execs are expected to accurately predict how much revenue they will sell. The past week of posts pertained to disconnects. 

Like sales and installs, we develop processes around forecasting disconnects. At Zayo, the disconnect forecasting process is in its infancy, though I suspect it is still more advanced than many other telecom companies. 

It is helpful to break disconnects into three categories:

1. Disconnects that might happen
2. Disconnect orders that have been received but the disconnect itself has not yet occured
3. Disconnects that have already occurred

A solid-as-a-rock disconnect process is one that enables management to focus on Category 1 situaitons.  If you have early warning of where you might be vulnerable, that probably means that your company is taking steps to avoid the disconnect.

Category 2 is important as well.  Often, customers will send in disconnect orders prior to their contract expiring.  Or, perhaps, there are ramifications, such as early termination penalities, associated with certain disconnects. A tight process flags these orders on the front end and ensures that the company is not short-changed in the process. As importantly, this stage sometimes provides the opportunity to make one last ditch effort to save the opportunity.

It is almost by definition that Category 3 means it is too late to save the order. However, Category 3 is important for forecasting cash flows–which BearOnBusiness.com readers know is one of my fundamental management principles. Moreover, accurate and timely reporting on Category 3 helps improve customer retention processes. 

We have a long way to go at Zayo before our processes will be what I want them to be. I’d like to see us using Salesforce.com as a tool for flagging potential disconnects in the same way we highlight potential sales. I’d like to see verification that we have spent weeks trying to avoid each and every disconnect before it hit our system. I’d like to witness the creativity and persistence of our employees as they turn churn opportunities into new long term contracts. 

We will get there, but it will take some time.

A title on one of my recent blogs was Like Fingernails on a Blackboard. One of the people on the Zayo team felt as passionately on that topic as me–but she felt a better title would have been Like a Pencil in the Eye. Well, I know she will be pleased that title was still available for this post.

“Disconnects just happen.”    

Give me a pencil. Give me Scarano’s eye. 

It is true that disconnects are sometimes unavoidable: a customer might move, their business might shut down or their needs might change in a way that your company simply can’t address. My educated guess is that the majority of disconnects do not fall into this unavoidable category. 

“I have a report that will show you are wrong,” you might be thinking. Be careful. Read the previous post. Is the report tabulated using ‘objective reality’ or is it influenced by ’subjective ass-covering’? Let me explain.

Let’s say we had a tracking process to explain reasons for disconnects. The process had only two check boxes.  

Check box #1 if the customer disconnected for any of the following reasons: 

• Customer moved to a geography ourside our service territory
• Customer went bankrupt
• Customer’s needs changed in a way that we were unable or unwilling to serve them

Check box #2 if the customer disconnected for one of the following reasons:

• I have no idea; I’ll check with the customer and get back to you
• I just got this account last quarter; can you check with the previous account exec?
• The decision-maker changed a few months back–I haven’t gotten around to building a relationship with her replacement yet
• I have bigger accounts I’m worried about
• Why would I care? My commission check is based on new sales. It is hardly worth my time to preserve existing revenue.
• The customer upgraded their service and, in the process, choose a different provider; by the time we found out what was happening, it was too late
• A competitor offered better pricing; if we only knew, we could have saved them
• The service from us was out-of-term. I didn’t want to bring this up for fear of losing them as a customer
• Our service has been spotty. If I were them, I would have switched too.
• The customer needed things from us that required a few simple modifications to our product. It is easier to move a mountain that to get something out of marketing and IT.

Got the picture? Assume you can do something to avoid the majority of disconnects. Use this assumption to improve customer retention. Don’t be too quick to rationalize that disconnects just happen.

Come to think of it, though, there are worse things in life than poking Scarano in the eye with a pencil.

Zayo Group is a roll-up of fiber-based telecom companies. We began (sort of) in late 2006. We researched every telecom company that had a fiber-based network. We contacted nearly all of them. We met with over half. We were impressed with some of the day-to-day operations. In other cases, we were simply perplexed. In any case, we were privy to the inner-workings of lots of the telecom boom survivors.

In some of the impressive situations, we observed a strong focus on preserving revenue. These teams either learned from the meltdown or, perhaps, they weren’t the ones guilty of ignoring their embedded base during the boom. 

Some situations were so-so. 

However, we also encountered companies that didn’t pay sufficient attention to disconnects. In a few cases, they didn’t even attempt to tabulate how much revenue was churning. 

Somewhat humorous were those situations where management teams insist their churn is low; yet, when pressed to provide evidence, they admit or discover they don’t actually track disconnects. ”But if we did track them,” they often persist, “You’d see how low they are.” Really? 

I’ve even heard, “We don’t track disconnects because they are low.” If there was a such thing as a ‘double really?’, I’d use it here.

Another situation we encountered is companies that had the data to support their claim of low churn. The data, however, was flawed. That is, the data was incomplete and/or inaccurate. 

Weak processes were often the culprit. Disconnects were happening–they just weren’t captured in those reports whose purpose was to tabulate disconnects. Obviously, this isn’t a good outcome. It happens frequently, as rigorous capturing of disconnects takes discipline, focus and tight processes.

Worse, though, are situations where management’s intent is questionable. Management, at times, becomes too enamored with their desire to show that disconnects are impressively low. That is, they don’t understand that the purpose of tracking disconnects is to provide a complete and accurate picture, regardless of whether the results are good or bad. Tracking must be the actual score of the game–not the desired score. The highest responsibility is for the reports to be thorough, accurate and objective. If the results are not good, address the problem; don’t blur the data. 

This sugarcoat tendency happens in various areas of business reporting–but it is likely more prevalent in the area of disconnects than in most other areas. Why? First, ascertaining disconnects and validating reporting is tedious. Second, it is highly unlikely the auditors, board of directors, shareholders or even CEO will tune into reporting methodology until problems become pronounced. By the time the problem is big, it is usually too late. The company now is in crisis mode. 

The first-order problem in crisis mode is what to do about the high disconnects and shrinking revenue. Once in crisis mode, the fact that inaccurate reporting contributed to the problem is, frankly, a secondary concern. Ironically, executives that thrive in crisis situations are in their element at this point, even if it was these same executives who caused the crisis by taking their eye off the ball.

Does your company track disconnects? Are the processes tight? Is the reporting accurate? Does your company use the data to measure its effectiveness in minimizing disconnects? If not, what will you do about it?

During the telecom meltdown, many companies discovered disconnects were a powerful force. Painful too.

The problem is that they discovered this DURING the meltdown (instead of during the disco revolution.) In many cases, the meltdown had already crippled their business.  Only then did they begin to focus on tightening their processes around disconnects; only then did they re-think their incentive schemes. By the time “disconnects” became part of the culture, it was a negative and devastating influence. In many cases, the last surviving memory of the boom-era customers is collapsing revenue.

At this point, you’re probably right–I am being a bit too dramatic. But perhaps not.

The overall theme of the BearOnBusiness.com blog is “what have we learned” as a result of the telecom boom and meltdown. Well, the need to rigorously and thoroughly manage disconnects is a lesson learned that many in our industry have yet to get a grip on.

This is amazing to me. I will repeat. Amazing.

How do I know this is true? See tomorrow’s post.

Do you like that as a tongue twister?

The past few posts referred to the disconnect inferno of the early 2000’s. “That was then,” is the hallway chatter I heard throughout the telecom universe. “It is good it is behind us.”

Is it? I don’t know if you noticed, but we might be smack in the middle of a recession. The telecom industry continues to consolidate. Technology continues to change. If disco returned, I might be somewhat nostalgic. If the telecom meltdown reared its ugly head, nauseated would be a better description how those in the telecom world would feel.

My point is this: disconnects need to be a focal point of operating a telecom business in both good times and bad. Preserving revenue is just as important–perhaps even more important–than selling additional revenue. Culture, business process and incentive plans should reflect this.

I hope those readers who are part of Zayo, Envysion and NGT reflect on this. In addition to observing evidence of this in your day-to-day work situation, I’d like you to ponder additional ways in which we can combat disconnects. And by the way, combat is a good word for this topic–especially in tough times. Remember the swimming upstream analogy–simply swimming harder is not a good plan. Combating disconnects requires focus, cleverness and persistence. More on this in future posts.

Visualize swimming upstream. As the current picks up, you will have a harder and harder time making progress. At some point, you begin to move backwards. So you try to swim harder. As you get tired, the current eventually overtakes you.

Disconnects are the Achilles’ heal in a recurring services business like Telecom. Disconnects are the current you are swimming against. If disconnects are high, it becomes increasingly difficult to grow top line revenue. If, in the face of high disconnects, you attempt to overcome them by selling more and more, you will eventually be overtaken by the Disconnect Inferno.

If you wonder what it felt like to be in the eye of the telecom meltdown, think about swimming upstream against a strong current. If you are still unsure, give a ring to Vonage employee and ask them how they are sleeping at night. Not well is my guess, as they are in the middle of a disconnect inferno of their own making.

Remember, bandwidth demand was still growing rapidly throughout the meltdown. Internet penetration was rising rapidly, consumers were switching from dial-up to broadband, and businesses were upgrading to DS3s and VoIP. Internet leaders such as YouTube, Myspace and Envysion were only then being invented.

The problem wasn’t lack of new bandwidth demand. The problem was how quickly the installed base was shrinking. The problem was the disconnect inferno. We should have seen this coming given the Trammps thinly-veiled message from the early 1970’s.

Picture yourself dressed in a white silk suit, blow drying your hair and dancing to a Bee Gees tune. It might have seemed hip at the time, but you knew there would be a price to pay. Though I was not quite a teenager, pictures of me in a sky blue leisure suit are floating around. Not fun.

Disco came crashing down around 1977. Exactly 25 years later, another bust occurred–telecom. I don’t know how KC and the Sunshine Band handled the end of disco, but the Telecom Meldown was painful for me and lots of my friends.

It is a little known fact that the disco craze and the telecom boom are inter-related. Like Nostradamus, the Trammps had visions–albeit blurry ones–of future events. It was one of these visions that inspired them to write their fantastic dance song “Disco Inferno”. As I explained in prior posts, Tina Turner caught on to this as did Madonna. For those keeping score, Tina and Madonna sold their telecom stock in 2000–it is what allowed Tina to divorce Ike and Madonna to buy a new home in Aspen, Colorado. Yes, what I am implying is true: if we were paying better attention during the disco era, the telecom meltdown might have been completely avoidable.

If you aren’t following me, I’ll give you a factoid that both Tina and Madonna knew. When the Trammps wrote their famous song, they took it to famed producer Tom Moulton, who mixed the record. He loved the song, but he felt the title was too long. Against the objections of the Trammps, Moulton demanded that ‘nnect’ be removed from the first word of the title. Henceforth, the song was known as Disco Inferno instead of its original name Disconnect Inferno. If you listen to older versions of the song, you can hear Jerry Mills Collins belt out the original “disconnect inferno” lyrics.

If you were in the telecom industry circa 2003, you know what it is like to be in the disco(nnect) inferno. Like leisure suits and blow dryers, not fun!

If you need a visual from the Saturday Night Fever era, click here.