Comments on Price Wars—High Speed Internet Wars

But first, if you haven’t seen our Pricing Cheers and Jeers–take a look here.

Also, for those of you in the pricing of services, especially for small companies, you should take a look at Ron Baker’s Versage Institute’s website.  Tends to have very good advice.  And, he’s written some very good books!  Check out the Pricing Manger’s Bookshelf.

Now to the day’s news.  Yesterday’s WSJ had an article by Vishesh Kumar: Price War Erupts for High-Speed Internet Service.  He talks about how Verizon is concerned because of a drop in their DSL users–due in part to some leaving for their FiOS service.  He also talks about how the market for high speed internet services is maturing quickly and that cable companies are picking up a bulk of the new subscribers.  Hence the price war as competing high speed services are trying to buy customers so they can sell other services to them.  This opens us up to a host of comments.

On the surface, the price war appears to be reasonable.  The market is “nearing” maturing”–which means it’s not there yet.  At 60%, penetration still has a healthy way to go and current growth of adoption probably confirms that.  So using price discounts to get new users is probably still reasonable.  Once you get a user using price, you know they’ll stick around for quite a while and you’ll be able to sell them a host of other services.  That is all except the “switchers” or true price buyers (yes there are even those guys in consumer markets) who will switch contracts when they can to take advantage of low price offers.  AT&T got bit by that several years ago in their DSL services when customers kept switching every time they or a competitor offered a lower price.  You can fix that by limiting the offers to non-prior users.  AT&T probably couldn’t track that with their current Customer Information Systems–ooops, probably lost a ton of money with something that could have been fixed by a relatively modest IT investment.

To look at the impact of price sensitivity, pricers and marketers can take the discussion at two levels.  The first one is anecdotal.  Me for example.  We used to use the Verizon service for cell phone, local, long distance, TV and internet but the service was lousy.  Lousy picture, couldn’t get unified billing, and expensive.  We moved all but the cell phone and internet to Direct TV satellite on price because the package included free installation.  The service was a bit better than Verizon but we started getting hit by price increases and nickle and dimed with fees.  When Comcast came up with a price promotion which bundled cable, local, long distance and high speed internet we bit for it and have been fairly happy–one bill, lower prices and better service.  And, we’re waiting for the Verizon FiOS high speed fiber optic system to be available at the house.  That’s a local contract issue with the town that Verizon is trying to solve at the state level. 

We are probably like a lot of other users (more on that in a sec).  It is highly unlikely (aka impossible) for us to switch back to Verizon’s DSL with price.  If they are losing customers to their FiOS–isn’t that a good thing for the company?  If we are unlikely to switch due to price, it means that price discounting is probably not the right tactic.  Two things are the right thing to do.  First, push like heck to get the local coverage for FiOS.  Second, be patient–until the service is available, knowledgeable customers aren’t going to switch back to DSL.  See, the anecdote can drive the strategy.

Anecdotal modeling like this works for most small to medium companies and for large companies when dealing with large customers.  You don’t need fancy modeling, you need to construct estimates of how big various populations are (users, non users, switchers, etc.) and what their likelyhood of switching is with various strategies, price discounts and others.  IT IS UNLIKELY THAT SOMEONE WHO HAS LEFT FOR QUALITY WILL COME BACK FOR PRICE!!!!!!!!  So don’t try to buy them back–fix the quality issues.  Don’t kid yourself (though lots of managers do).  Sit managers in a room and estimate how many customers and how much revenue you have for customers in various “states”.  By states I mean: happy users, at risk users, switching users, competitive users (for each competitive) with the same levels, and non users.  Yes there is more depth needed here but you should get the point.  Those estimates can then for the basis for a more informed development of strategy.  Test the strategy on your estimates and that will tell you whether it is the right thing to do. 

We did some work recently in the medical business where lots of the senior managers wanted to use a penetration price for a new product.  Using a “scenario analysis” where internal knowledge was used to build the states and market picture, it was clear that a neutral/skim strategy was more appropriate.  The product was introduced, prices held, no price war developed and the company got more share than they anticipated–they had a terrific sales force that knew how to sell value.  This stuff does work if you do your homework.

If you are a larger company with the resources to understand your markets, rather than dumping all your money on fancy segmentation studies or demand reserach that tries to figure out market responsiveness to prices, try doing something different.  Spend your money determining a) what the possible customer states are, b) the size in quantity of customers and revenue for each state and c) the sensitivity of people or companies in each state to changes in strategy–price discounts or otherwise.  This is what some conjoint consultants do but they don’t model all the states or all the possible changes in strategy–they focus on price changes.  I would expect that some have begun improving on this and would love to hear who is.

This is call markovian or stochastic modeling.  Also known as lilly pad analysis (frogs jumping to different lilly pads on a pond), it is a terrific way to get a handle on the proper strategic approach for new and existing products.  You can estimate the states with internal experts or you can do market research to confirm or do the estimates.   And, it will tell you when pricing will help, when it won’t, and when you’ve got to be working on other elements of the mix instead.  Know of another medical device that was trying to buy their way into the market but costs were too high.  Instead, they needed to focus on high value users where price wasn’t an issue.

The Verizon example is a great case where markovian analysis would shed a brighter light on the more appropriate strategic approach—one that would probably not include lower prices.