EXPERT ADVICE

The Top 5 To-Dos Before Kicking Off Your E-Commerce Project

So you are ready to embark on an e-commerce project. You are not alone.

The Economist found that of 15 major industries studied in late 2008, only one — e-commerce — had a sunny outlook for 2009. My analyst (techno, not psycho) tells me that 20 percent to 30 percent of firms that do business online will start a major e-commerce project in the next twelve months. Whether it is despite the current economic situation or because of it, many firms have plans for multiple e-commerce projects over the next few years.

A typical e-commerce project has high, obvious financial and branding impacts, which can make you a star if you pull it off. That’s the good news.

The other news is that e-commerce projects are nothing if not visible. Your customers, executives, competitors, colleagues, cousins and dental hygienist will all see your work and have an opinion. You are like the manager of the Yankees. The stakes are high.

We’ve basked in the glory and sifted through the wreckage of a decade’s worth of e-commerce projects to bring you this list of five things to help stack the odds in your favor:

  1. Have the guts to proclaim specific goals.
  2. Make the team a Team.
  3. Go agile — it’s not just for IT anymore.
  4. Involve customers.
  5. Don’t reinvent the wheel.

Have the Guts to Proclaim Specific Goals

It’s amazing how often we see e-commerce project teams with little inkling as to why they exist. I don’t mean this at a metaphysical level. A simple goal or two will solve the problem. Here are a few examples:

  • Raise conversion rate from 1.8 percent to 2.2 percent in a year.
  • Double traffic in the next two years.
  • Enable business users to change promotions in an hour instead of a week.
  • Add site visits at twice the rate of our competitors’ average.
  • Launch new micro-sites in three months for under $300,000 each.

As important as the specificity of this message is the need to have the leadership voice it, starting with the executive sponsors. John F. Kennedy’s 1961 “man to the moon” speech provides a good template.

The practical side of this is that when the time comes to decide whether the next 300-hour change order goes to task A or task B, you will need criteria. A quantitative list of posted goals for the project will help.

Make the team a Team

Here comes the touchy-feely stuff. Unfortunately, even to a grizzled veteran like me, this is crucial. When you are in a trench, you are fighting for the guy next to you, not for some general. If you like the guy next to you, you may fight harder than if you didn’t. The challenge with e-commerce projects is that you have Martians and Venusians — IT and marketing — in the same trench.

You can’t legislate culture, but you can give people an opportunity to get to know each other. Run fun social events regularly, and start early. We’ve gone go-karting, rock climbing, curling, hiking, on pub nights, and in charity runs, just for a start. When the pressure hits later on, this can be the difference between a concerted team effort and violence. Crasser team-building tactics such as financial incentives can help as well, but there is no substitute for interpersonal relationships.

Go Agile – It’s Not Just for IT Anymore

For the uninitiated, Agile is a software development methodology that emphasizes flexibility.

Agile enthusiasts will say it never was “just for IT.” That aside, an Agile approach addresses one of the key difficulties of e-commerce projects — continuous, externally driven change during the project period.

Over the course of a six-month project, your competitors will change tactics many times, your priorities will shift, and you will learn a lot about the capabilities of your technology and team. Why try to set things in stone at the start when you are at the apex of your ignorance?

Agile is about taking the pain early. By forcing yourself to test a set of complete functions — end to end and at very regular intervals — you discover the bad news while there is time to fix it. You must put a lot of pins in place to have this capability, including:

  • Regression tests that employ a lot of automation and can complete quickly;
  • Ongoing test case management;
  • A tolerance for ambiguity; and
  • Strong leadership.

Agility also drives team structure. Rather than having teams with names like “front end” and “database” and “storefront,” consider teams organized around workflows or complete functional areas. For example, the “promotions” sub-team might consist of people with skills in marketing, customer experience, HTML, Ajax, Java, systems management and testing — everything needed to make “promotions” meet their business objectives.

Like Einstein’s Relativity theory, Agile is simple, but it’s not easy. You need to train everyone, ideally in mixed classes (IT, Marketing, etc.), so they can see each other’s perspectives. You need leaders committed to prioritization over certainty, and a steady, slow but complete roll-out of functionality rather than sexy early demos followed by panic.

If you must use a Waterfall approach, at least set 45-day milestones where you will completely regression test the system, have end user/customer walkthroughs, and run it through its RASP paces.

Involve Customers

As a young, cynical techie, my friends and I liked to say “this would be a great job if not for our customers and sales reps” (the humor was in the irony). By the level of customer involvement in some e-commerce projects, there may still be people out there who think that way.

Early in the project, when you have basic workflows and wireframes done, you will want to put them in front of a sampling of the people who will use the end product. Of course, “customers” may include customer service representatives, suppliers, the general public, marketing staff, and so on.

I would suggest using more than one approach. Focus groups can give you in-depth insights but may not be representative. Secondary research, such as correlating Gomez and IR500 stats to your competitors’ tactics and results are good “virtual” tests of customer acceptance. None of this is tremendously time-consuming, but it is very high-leverage, and it will help you build the A/B or multi-variant test plan that is critical to early post-launch tuning.

Don’t Reinvent the Wheel

If you’ve done #1 and #4 well, then you should have a good idea of where the leverage points are in your project. That should guide your Build vs. Buy vs. Leverage business cases at a functional level.

We’ve seen projects where business analysts spent many hours in requirements definition on features that were secondary to the overall goals and largely provided out-of-the-box in their e-commerce platform.

Thorough training of business analysts in two areas will help prevent these very expensive errors:

  • What is available in your current set of packaged software, including in the vendor’s near-term roadmap? and,
  • What is available in the market, off-the-shelf?

In Summary

Here’s your quick checklist. None of these need to cost a lot of time or money, but from our experience, they will help you cross the finish line with a smile. Good luck!

  • Post specific, quantitative project objectives, signed by the executives.
  • Schedule monthly social events that are fun and don’t need “selling.”
  • Run Agile training courses with mixed classes.
  • Run a focus group and perform secondary research, then publish it to the whole team.
  • Train business analysts on out-of-box functionality in current and potential packaged offerings.


Gord Janzen is chief operating officer of Elastic Path Software. He’s responsible for strategic planning, professional services engagements, and the day-to-day operations of Elastic Path.


1 Comment

  • It’s a shame that with all this, you still only have half a market, at the most.

    No mention of "Localization", which could well double your market footprint. Fear of other languages & cultures is similar to fear of investment or fear of digital tech, but like the latter two, economical solutions for professional localization are out there.

Leave a Comment

Please sign in to post or reply to a comment. New users create a free account.

Related Stories

LinuxInsider Channels