Ten Ways to Fail at E-Commerce
- 20 July, 2000 12:01
TORONTO (07/19/2000) - Here's fact one about e-commerce in the year 2000: nobody - nobody - has a reliable prescription for success. Not even us. It's just too early to say, "This works, guaranteed." Don't let anybody tell you otherwise.
That said, e-commerce consultant Carolyn Burke may have a sure-fire recipe for failure. Burke, president of Toronto-based FSC Internet Corp., recently shared with us her Late Night With David Letterman-style "Top Ten Ways to Make Sure Your E-commerce Project Fails."
They're guaranteed to bring a smile to anyone who has ventured into the turbulent and fast-moving waters of e-commerce. But when we talked to Burke, we also discovered a depth of real-world experience and advice behind her tongue-in-cheek barbs.
RIGHT NOW, ON OUR SHOW
We'll get to both barbs and advice, but first, as Dave himself might say, we want you to meet our next guest.
Without necessarily meaning to, veteran consultant Bob Fabian, now director of the Seneca College Internet Commerce and Technology Institute in Toronto, puts Burke's recipe for e-commerce disaster into context. Fabian is even more skeptical than we are of prescriptives - positive or negative.
"Every time I probe the obvious questions about e-commerce and ask, 'Are there absolutes yet?', the answer is still 'Probably not'," Fabian says. "Absolutes assume we know how the story's going to turn out. We don't."
To succeed at e-commerce, Fabian believes, you need to find a balance point in several dimensions. And the fulcrum will be different for every organization.
For example, you certainly need a well-thought-out and workable e-commerce "value proposition" - a mix of product, price, ancillary services and so on that is sufficiently differentiated from competitors' and attractive to the market. (This, sadly, is not as self-evident as it might seem, Fabian notes.) But at the same time you can't be too rigid about your business plan.
"If you just throw up a Web site and hope people flock to it - without ever thinking about the business idea - it's going to be flabby and unfocused," he says. "The flip side is companies that totally ignore the market and keep on trying to do it anyway, when the market is saying, 'We're not interested in this'."
In the same vein, you have to find the right balance between planning for the future - what happens if your site is an overnight success, for example; can you muster the resources to handle the crush? - and staying open to the zig-zagging vagaries of the Internet market.
"You don't just charge forward with no thought for tomorrow," Fabian cautions.
"But by the same token, crossing every 't' and dotting every 'i' is going to be impossible given how fast you have to respond in this market."
Likewise, on the crucial question of outsourcing, companies need to find a comfortable mid-point. Neither of the extremes - outsourcing everything, taking advice from no one - is workable. The former means you potentially put your whole future in the hands of outsiders. The latter is patently foolish. But each company, Fabian says, has to figure out the right balance.
While the professorial Fabian - he's in fact Dr. Robert Fabian - is inclined to take the high-level view, Burke, an in-the-trenches e-commerce veteran, focuses on getting the job done. And occasionally on failing to get the job done. As you explore her top ten ways to fail at e-commerce, keep Fabian's agnosticism in mind. We'll help by occasionally injecting Fabian-like counterpoints.
For added perspective, we asked FSC customer Greg Young, director of corporate public relations at Mazda Canada Inc. in Toronto, to provide a client-side commentary on Burke's top ten. Mazda has been working with FSC since 1996 on a splashy site that stops short, for obvious reasons, of selling cars directly on-line, but is in other ways exemplary.
TOP TEN WAYS TO FAIL
10: Hire several different groups of consultants, and make sure you don't give them clearly defined areas of responsibility. That way you can be sure they'll wage a turf war instead of carrying out your project plan.
Burke agrees that most companies need to bring in outside expertise - no big surprise since her company is one of the candidates. But selecting the right outside experts is crucial, she says.
Burke tells of a firm that hired one consulting organization to handle the technical side of its e-commerce initiative, but also brought in another consultant to manage the overall project. That person in turn brought in other consultants.
The project manager knew little about Internet projects but would not take recommendations from the experts she and the client had hired. Result: the site couldn't meet performance objectives, or the project its deadlines.
"If you find your outside people are getting into any kind of turf war, it's important to weed out the trouble-makers as fast as possible," Burke says.
To prevent it happening in the first place, carefully screen consultants - not just for the right expertise, but also to make sure they have a history of working well in a team. It means diligently checking references.
"If I was pulling together a team, I would do that," Burke says. "If it hadn't been done, I'd very quietly mention to the client that they have a problem and suggest what they could do about it."
Fabianistic, yes, but: Hiring multiple consultants - big companies do it with ad agencies, for example - might hedge your bets in some cases. It might also encourage healthy competition that ends up producing better results.
9: Be sure you don't budget for the project in your annual plan. This ensures that you have to scavenge for resources, and that you have to choose all the cheapest possible options, whether or not they're any good.
Sometimes the e-commerce mandate - and budget - come down from above, but sometimes the initiative comes from middle management. If project champions can't get senior management buy-in, they may try to wing it on existing budgets. That's when problems can occur, Burke says.
A penny-ante, scrounged budget often means servers without the capacity, or links to the Internet backbone without the bandwidth, to handle anticipated volumes. Or half-measures in the design and execution of the site, which inevitably turn customers off - only a few products listed at the site, for example, or no mechanism for actually buying on-line.
The other risk is budgetary turf wars among departments involved. "Unless it's clearly spelled out up front who is contributing what and who has what control, you're likely to have turf wars later on," she says.
Burke knows one large manufacturing company that designated managers of 12 product areas to participate in an e-commerce project. Each had to contribute a share of the budget. Some way into the project it was decided the site would include products from only four of them.
"It was awful," Burke says. "There was no cooperation [among the product managers] after that decision was made. The site never launched."
Mazda Canada, notes Young, meets with Burke's team every year to establish an annual budget for ongoing development of the company's Web site.
Fabianistic, yes, but: If a great e-commerce opportunity pops up mid-year, do you ignore it?
8: Ensure that your advertising and marketing people can't stand your technical people, and vice versa. (Using groups who've never worked on an e-commerce project before is a good way of accomplishing this.) The result will be things that look good and don't work, and other things that work but look like a dog's breakfast - both highly desirable from a failure point of view.
Traditional IT departments are there to support the technology infrastructure and don't really get marketing. And extroverted marketing types, more used to thinking about brands and products and pricing, know as little about IT. But now they're strange bedfellows on the e-commerce project.
"Unless you have a translator," Burke says, "the project is going to fail. You really need somebody to pull those groups together." That's where companies like hers can make a contribution, Burke contends.
Young admits there have occasionally been disconnects between IT and marketing at Mazda, but with help from FSC the company has always worked past them.
The results of not pulling the two groups together, Burke says, are often comical in their awfulness. As a form of entertainment, she and colleagues exchange URLs of sites obviously created by IT nerds. Blue text on a white screen, blinking text and no coherent design theme are unmistakable traits, she says.
7: Bring in an outside project manager who understands neither your business, nor the ins and outs of e-commerce projects. This will help you to get the misguided project planning that is essential to a truly resounding failure.
This is a familiar failing among dot-com start-ups, Burke says. Typically launched by bright young techies who lack business skills and project experience, they inevitably go outside for help. Too often they pick a person with experience in other start-ups. "And nine times out of ten that means it's someone with experience of failure."
One strategy for dot-coms, she suggests, is taking a recommendation from a funding venture capitalist. An even better solution is for one of the founding partners to learn the business side and take on the role of directing the e-commerce project. "They know the idea and have the enthusiasm and energy," Burke observes.
In fact, on any e-commerce project, it's better for overall project direction to come from inside. "Usually we would expect an organization to have a director of e-commerce, someone who understands the company's business needs," Burke says. "We don't typically go in playing that highest-level role."
That's certainly the case at Mazda, where Young has overall responsibility for the e-commerce site while relying on FSC for technical and graphic design and recommendations on e-commerce strategy.
6: Try to do the project entirely with existing staff resources. This will guarantee a healthy level of incompetence, which will generate the lengthy delays a truly massive failure needs!
This is the one point of Burke's that makes Young blush a little. While he lets FSC look after the technology and design, his objective has been to get as many people inside Mazda involved in the project as possible.
"I believe if we're really going to integrate e-commerce into our operations, it has to be part of everybody's job descriptions," Young says.
So the people responsible for product planning - keeping dealers apprised of what's coming next - need to be involved. The marketing communications people who work on brochures should help make sure the Web site is up to date. Ditto for sales and marketing on pricing issues.
Burke is really more concerned about companies that rely entirely on in-house marketing communications and IT staff to drive a project. These people typically have no experience with e-commerce or feeling for the new medium.
Plus, they usually have other responsibilities that demand their energy and attention.
"Generally it results in a Web site project that doesn't have a deadline," Burke says. "It gets tossed up there whenever it's done. No one likes it. The navigation is confusing. It's probably very slow to display pages - all the attributes of a bad site. We often get called in after that kind of project."
5: Don't let anybody else in your company know what you're doing. That way, if your project starts to succeed, they won't be ready to support it.
One common mistake is to leave sales out of the loop. "If the project budget comes out of IT, the last thing they want is to share it with sales," Burke notes.
Sales, being externally focused, tends not to pay much attention to infrastructure. It's more likely to wait until a project is complete and then say what it doesn't like - such as the fact the Web site is cannibalizing its own sales. "It's important to bypass all that and get them involved right up front," Burke says.
She also knows one dot-com where not adequately involving shipping and finance created chaos. It was company policy that Web credit card transactions were only completed when the ordered product shipped. Faced with sudden pressure from above to process Web orders ASAP - to drive the payment process - shippers began sending whatever was available, whether it was what the customer ordered or not. Result: a huge increase in costly returns.
"If you cut out any one department in a large organization," Burke says, "you create a weakness in the business plan." Self-evident? You would think so, but apparently not.
4: Use all the latest buzzword-compliant technologies, whether or not anybody has actually got them to work.
In infrastructure areas such as Web servers and Internet security, there are vendors with lengthy track records and solid reputations, and then there are a lot of wannabes, Burke says. The wannabes may have great new bells and whistles or a better price. Trouble is, nobody has ever actually used their products in the real world.
This doesn't mean you can afford to ignore innovative new products and companies that could stake you to a competitive edge. "If there's some new, exciting software solution, take advantage of it by all means," Burke says.
"But make sure you're picking a vendor that'll be around next year - or next month."
Motherhood and apple pie, perhaps, but Burke says she regularly sees e-commerce projects stumble on just such lack of attention to due diligence.
Then, of course, there are the well-established vendors with a lengthy track record of consistently poor performance. E-commerce companies buy it anyway.
3: Use whatever products your existing vendors have, rather than best-of-breed solutions. After all, what's a little failure between friends?
Burke worked on one project where the company made a top-down decision to use the firewall hardware and software of an existing supplier. The firewall had been deployed in the hundreds or thousands, but had never operated in the real world at the scale this company needed.
"It failed dramatically," Burke says. "It couldn't handle the volume. They hadn't done their homework. They spent a lot of money and time, and they lost it."
The trouble is, everybody wants to be part of the e-commerce bonanza. Burke knows companies that built their reputations in areas almost completely divorced from the Internet, but still have e-commerce offerings they sell into their installed base, usually untested.
What she says makes sense, but what happens if the best-of-breed point solution turns out to be incompatible with the company's legacy systems? Hmmm.
2: Don't bother to check out what your competitors have done. If you learned something from their successes, that might compromise your ability to fail!
Dot-coms are prone to a corollary of this rule - assuming their latest brilliant idea is totally original, keeping it under tight wraps, and then finding when they launch that others had the same great notion. Then their e-commerce business turns into something quite unexpected and possibly unfeasible: a dog fight.
For aspiring clicks-and-mortar companies too, checking out what competitors are doing may tell as much about what you shouldn't be doing as what you should.
"If my competitor is losing money going on-line and losing it in a very public way, do I really want to go ahead and lose money too on the chance I'll keep some market share?" Burke asks. "If I was selling books, the last thing I'd do right now is go on-line. The most I'd be doing is defending my market position.
In business you want to grow your market."
Young says he regularly benchmarks competitors' sites. In recent months, he's seen some competitors - Ford, Toyota, Volkswagen - begin to experiment with direct on-line sales. But Mazda will likely never go that route in the interests of keeping dealers onside, he says. If the market demands it, the company might look at automating the matching of site visitors with local dealers.
Mazda has also, quite deliberately, not gone the tons-of-data-about-each-car route followed by most competitors, preferring to differentiate itself by providing interactive components such as a payment calculator (now mimicked by every other vendor), and by concentrating on sizzle, making the Web site mirror the company's distinctive TV campaign.
1: Keep adding features prior to the launch of your project. If you never launch, you never run the risk of succeeding!
Sometimes companies are unwilling to launch their site until they deem everything to be up-to-date and perfect. The site may be ready to go but the company keeps adding more content.
"We would draw a line in the sand and say, 'Any new content will be Phase 2; we'll go with whatever we've got there now'," says Burke. Without such commitment, the project may fall months behind. Or it may never get off the ground at all.
Many new e-commerce companies don't understand the need to move forward decisively, because they've had little experience of the traditional custom software development process.
"The software goes through a process of being designed and prototyped. Then you build and test the code, then you launch the code. Then you start on version two," Burke explains. "An e-commerce site should be the same."
No quibbles from us on that one.
FLIP THE PENCIL, DAVE
There they are: the top ten ways to fail at e-commerce, a list with wit and wisdom for e-commerce managers everywhere. Just keep in mind, we're still at the dawn of the e-business era. Nothing is written in stone yet, not even these rules - which Burke herself realizes. There can't be anything more transient than a Letterman-style top ten list.
Tony Martell is a freelance writer specializing in information technology and IT management. He is based in London, Ont.