This blog entry was posted by Ed Margulies, Chief Operating Officer of FACE IT Corp. ("Putting a Pretty Face on Customer Service"). Margulies is a telecommunications architect, usability expert, inventor, and the author of 17 books on telecommunications, contact centers and service automation.
Crash, Burn, Shame, Flames
That's how Ted Tencza, a dear old colleague from Dialogic, used to describe poorly planned services that went down the tubes. Over the years, we've witnessed many ill-fated services fail. And they run the gamut from dial-a-date to virtual contact centers to voice messaging service bureaus. Technically, most of them worked OK, but there are critical planning mistakes to learn from.
Fee Collection
Nothing kills a decent CaaS-based offering better than a lack of cash flow. You need cash to make payroll, build new product, and keep the lights on. Unfortunately, many entrepreneurs take it for granted that the people who are benefitting from the service are actually paying for it. There are a few approaches over the years that have worked to varying degrees.
For example, in the heyday of 976 dial-it services, the formula was simple: You put up a service, the phone company collected and they paid you your cut as an information provider. In the beginning, that was a perfect alliance, because if a caller didn't pay his 976 bill, his phone service got cut off. That was of course until the phone companies lost their guts and eventually allowed users to selectively pay portions of their phone bill. The point is, one way of collecting fees is to let a partner who has the infrastructure and wherewithal to handle that do it for you.
Another model is to use sponsorship as a means of subsidizing the operation. This was pioneered in the 1930s by the Audichron company who sought sponsorships from local banks who subsidized the "talking clock" phone service they offered. In essence, the cost to the consumer is nothing and you as the service provider are being paid by a corporate sponsor.
This sponsorship paradigm also extends to advertising schemes where use of the system is subsidized by corporate sponsors who pay for the placement of electronic ads associated with the use of your service.
With the advent of internet-based transaction processing services such as PayPal, it is possible to bill and collect for services from users via these third parties.
Regardless of which approach or approaches you choose, your fee collection plan must be well thought out. Think through: 1) What is the cost of administering the plan?; 2) If it is based on advertising or sponsorship, what is the cost of hiring a sales force to sell the ads and what is the spin-up time for revenues to start flowing?: and 3) How reliable is the mechanism you have chosen and are there back-ups to the plan?
Underestimating the need for Support
I speak to many entrepreneurs and service providers who think Wikis and self-service knowledge bases are all you need to take care of customers. I speak from experience: that may be true in the beginning, but once you reach a critical mass of tens of thousands of users, it just begins to fall apart. Why? Because no matter how hard you try to dodge the bullet of doing some kind of support staffing, customers who demand service will always find you, the owner, and press you into service. And why do you cooperate? The threat of bad press. Today, with so many users hooked up to YouTube and social networking sites, it only takes 140 characters to put the message out there that you suck. Remember the YouTube video about the Comcast employee who fell asleep on a customer's couch waiting for service center staffers to answer the phone? Even Amazon thought they could get away with no contact center in the beginning.
Besides hiring a support staff and building-out a contact center, there are other ways you can provide support. For example, you can outsource support lines and use services like GetSatisfaction.com. Here, a third party provides you with widgets for your web site that link to threaded discussion boards and electronic support forms that make it easier for you to answer questions customers pose. Be sure to check out these services as part of your own customer service planning so you can anticipate the true cost of providing support.
Even contact center infrastructure can be rented now. As an alternative to buying your own ACD, IVR, etc. you can timeshare network-based services. That sure beats building-out your own contact center. But the street price per agent is still around $100. Companies like NexxPhase (subsidiary of Atlanta's NexxLinx), beyondmortar, and Arise offer virtual contact center services you can check out.
Lack of Promotion
Only in movies like "Field of Dreams" will people come just because you built it. Great movie but a bad idea on which to build a real business. People will not magically show up because you exist. You need a plan to get the word out. How can you get the word out? There are many ways. First, it's a good idea to tap in to the power of the analyst community for your vertical. That might be Gartner, or AMR, or Opus Research. It is important for industry analysts to understand what you are doing and to even influence your launch plans. Making a disclosure to analysts pre-launch is critical. Why? Because analysts provide the press with meaningful context, quotes, and sidebars that can be published. Otherwise, when you run to the press with no analyst back-up you just make it harder on the writer. Making anything hard for the writer means you get short shrift. Don't make the mistake of going to the press without having spoken to and worked with the analyst community first.
If you have your own in-house resources for public relations, that is good. They can review editorial calendars, conference call for papers, and schmooze the press. If you don't have these resources, you need to hire a PR firm. It's important also to make sure your press releases are actionable. That is, embed an offer for a white paper, glossary, free download, and the like so the magazine or web site reporting can create a sidebar with that offer.
Viral marketing and the social networking scene also provide a good outlet for getting the word out, but you need to establish a persona and stick to it. For example, if you want to broadcast on Twitter, create the persona and send out microblogs frequently. And the tweets need to point to something. Guide people to point to pictures, web sites, coupons, offers, blogs, and so on to keep interest. You also need to establish a dialog with your followers and encourage them to ask questions and respond to your tweets. This is a lot of work. I know because I tweet on behalf of an animal rescue (non-profit charity) and getting up over 1,000 followers this past year was a chore.
Of course there are traditional ways to promote such as advertizing. On-line advertising nets over $25 billion annually so it must be working. But just throwing ads out there can be costly and wasteful. Make sure there is some call to action. You want people to respond to a competitive challenge, a contest, a beta challenge - something. You can even take an educational angle and offer glossaries, white papers and technical data. The point is, there must be actionable items in your ads. Be sure to collect metrics on hits, conversion, and overall efficacy of your ads.
Rapid Response to Feature Requests
A big failing of numerous services is the inability to roll with the punches from a feature demand standpoint. It's OK to start with something useful but small and make incremental improvements. But users are fickle. If you don't keep pace with their demands, they will find another provider.
What does this mean from a planning perspective? It means you either have to do the programming yourself, or you have to hire someone to do it for you. Practically speaking, this means you need to maintain a good relationship with your programming staff - regardless of their status as full time employees or contractors. Nothing kills a service faster than it being deemed "weak" by the user base.
Dealing with Resiliency and Downtime
I don't care how good your software is. There are a lot of moving parts. And services have to plan for downtime, upgrades and resiliency. Many times, you will have a service outage not because of "bad software" but because you failed to plan for resiliency. For example, if your web site is hosted and your customers use it as part of your service, they must have unfettered access to it. If you are configured for only one server with no redundancy and no load balancing, you are dead in the water if something goes wrong with that server.
From a planning perspective, this means you need to contemplate paying extra for redundancy, multiple data centers, load balancing, battery back-up, etc. If you plan on using a hosting provider for your web and telecommunications infrastructure, don't take anything for granted. Visit the data center. Check out the battery backup. Ask to see the multiple servers, load balancers and switches.
If you decide to self-host, be ready for some pretty big expenses. You'll want redundant LANs, dual firewalls, redundant NICs, redundant entrance facilities, backup servers, etc. Either that or deal with being "down" because everything is crammed on to one or just a few servers that have no fail-over or load balancing.
It's great to have a dollar and a dream. Many wonderful and profitable services have in fact been launched on very little money. But if no one knows you are there, there's no money. And if you have not figured out how to collect money, you'll go out of business. And if you don't provide adequate support, robust features, and reliability, your customers will force you out of business and move over to a competitor. Address these top five issues from the start and you will avoid the common pitfalls repeated time and again by would-be service moguls.
Technorati
Del.icio.us
Slashdot
Digg
Recent Comments