I visited Forrester’s headquarters in Cambridge, Mass. last week and faced the difficult task of finding hotel accommodation. High demand had obviously forced prices up, so my usual preferred properties were charging double or triple what I was willing to pay. I used three different websites to try and find a suitable alternative — two hotel chains and one aggregator — but was left seriously disappointed, if not surprised, by how useless they were. The sites could support neither my digital customer journey nor my physical one.

As an analyst who studies eProcurement (which is B2B eCommerce from the customer side), I’m intolerant of bad eCommerce site design and execution — it angers and offends me. Having also studied outsourced development, I can guess not only how these websites are deficient but also why. These companies have spent huge sums on developing customer-facing software, and yet they’ve utterly failed to meet our needs as business travelers. Worse, they’re failing to iterate, to learn from and fix quickly their products’ failings. Here are four lessons that everyone can learn from these bad examples:

  • Help business shoppers balance price with value. Value to the business traveler is a combination of many factors, including hotel quality and travel time to the target location. We’re willing to pay a small premium (within the limits of our company’s policy and what we can sneak past our manager) for higher value, but we can’t ignore price completely. The hotel sites all fail to understand this tradeoff. You can sort by price or quality or distance, but that’s useless. The sites could all suggest cheap hotels that were over an hour’s drive from the office or nearby hotels that were ridiculously expensive, whereas what I wanted was something in between. The sites offer filters, but these don’t solve the problem because the tradeoff decision isn’t binary. For instance, a filter by price and distance might exclude the best-value property that was 0.1 mile further away and $1 more expensive.
  • Understand what value means to the individual so you can personalize recommendations. Value is a combination of several factors — such as quality, diverse supplier, delivery date, green item, etc. — that each person will weigh differently. For instance, some business travelers will go the extra mile if there is a gym on site, whereas others will pay extra to avoid a commute. In this context, the hotel sites’ filter by distance is misguided; it’s travel time and cost that matter. I had a rental car, which gives me more flexibility but also means that I want parking charges included in the hotel cost. Time-wise, 15 miles out of town from the office on Route 2 is equivalent to two miles in the opposite direction. Without a car, a hotel close to a Red Line station would be fine, but otherwise I’ll have to factor a taxi’s cost into the price/value comparison. Why can’t a booking site work this all out for me?
  • Developers should also be users. I bet that one of the reasons for these deficient sites is that neither the hotel chain’s employees nor its offshore developers have experience of shopping around for a hotel like this. For instance, if any of the team were frequent travelers, they’d know that a search for three nights means “for each night,” not “for all three nights.” The sites told me, incorrectly, that nothing was available within my parameters when in fact there were several options if I would tolerate the mild inconvenience of switching hotels for the third night. It’s a common complaint from my research sources that their offshore development service providers’ staff lacks the contextual experience to be able to contribute effectively to co-creation. It’s hard to build great software for western consumers of financial services, travel, healthcare, luxury goods, etc. if you’ve never bought or used them yourself. Ride-alongs and usability testing can only do so much to bridge this contextual gap. Development leaders who are experiencing this problem should consider using local service providers to augment offshore teams, even if their day rates are higher.
  • Monitor usage and react quickly to obvious problems. One of the hotel sites I used recently suffered a major reengineering recently as part of a corporate rebranding program. The initial release was unusable, in my opinion, in that it lacked several basic features, such as listing available hotels in my search zone sorted by price. That’s bad, but underestimating what MVP means is forgivable. What is unforgivable is failing to fix those flaws within a few weeks of the initial release. I was shocked when I revisited the site a month later to find that it hadn’t improved significantly. That risks losing my loyalty forever. Savvy business users know how Agile works, so we tolerate imperfect early versions provided our suppliers rapidly iterate to fix design flaws, which should be obvious to them from analysis of our site traffic. A simple FeedbackNow button on the site would be even better — loyal travelers would have been delighted to rate the new site and suggest improvements.

Bottom line: This is a lesson not only for hotel chains but for all B2B eCommerce leaders. You can earn customers’ loyalty by providing better personalized recommendations that understand their individual price/value tradeoff. That involves a nuanced weighting of multiple factors, not binary on/off filters nor a choice between equally useless sort orders.