I’ve been following Dharmesh Shah’s OnStartups blog for years and I remember when he announced HubSpot, the company he was starting. I’ve been fascinated to watch it grow and grow, so I was excited when I got to visit their offices a few months ago. Just after my visit they closed a Series D funding round for $32 million from Sequoia, Google Ventures and Salesforce.com, but despite its success almost nobody in the technology world has heard of HubSpot. I blame the combination of a location in Boston and a mainstream customer-base of small business owners for the lack of recognition. It’s a shame because there’s a lot to learn from their technology and process — they’ve solved some hard problems in thought-provoking ways.
People are fascinated by mirrors
There’s a good chance you’ve used their Twitter Grader tool, and its popularity shows one of the secrets to HubSpot’s success. The inspiration for the company came when Dharmesh realized that his own blog was driving a lot of traffic, and the startups he was helping out were all struggling to get anywhere near the same number of visitors. He built HubSpot by applying what he’d learned from blogging, and one of the key lessons was that people crave new information about their own lives and projects. If you can create a service that gives people interesting data about themselves and their organizations, they’ll spend time exploring it and they’ll share it with their friends.
With Twitter Grader, Dharmesh didn’t just create a source of free advertising for his company, it’s also implicitly targeted at people who want to improve their presence on the social network. Many of these people will be the small business owners that are in his target market. Even better, by offering the statistics as a gift to users, he created a small sense of reciprocal obligation that will make them more likely to purchase his services. The approach started with their original Website Grader service, but they found it so powerful, HubSpot now has a whole range of similar free tools for analyzing everything from your Facebook page to your blog.
The lesson for me is that giving people data and visualizations about things they truly care about can be a powerful tool for drawing them in to your service. Do some creative thinking about your customer’s problems, and see if there’s something you can offer them as a reward for their attention.
You should kill unicorns and rainbows with science
One of the most enjoyable conversations I had at HubSpot was with Dan Zarrella, who describes himself as a social media scientist. I can already hear some physics PhDs grinding their teeth, but Dan has earned that title by applying a lot of much-needed rigor to the fluffy world of social media measurements. He’s crusading against “unicorns and rainbows” metrics that have no connection to the goals you want to achieve. Many businesses have focused on building up easy-to-measure numbers like fan or follower counts, but to use Eric Ries’ term, those are just vanity metrics. You can gain a million friends without it leading to a penny in revenue.
Dan’s antidote is the relentless application of logic and analysis, working backwards from the business goals to evaluate everything you’re doing as objectively as possible. A fantastic example of this is his study looking at how minor content details, like punctuation, make a retweet more or less likely. It’s possible to argue with particular conclusions he draws, but he’s transparently laid out the methods by which he arrived at them. Anybody with some technical knowledge and access to a decent chunk of Twitter data can try to reproduce and refine his results. This makes the report so much more useful than the opinions or impressions that dominate most discussions of social media, since we can actually have an evidence-based argument about it.
I came away from talking with Dan with a new appreciation of how powerful the scientific method can be in even the most unlikely situations. I’ll be taking a fresh look at some of the painful problems my projects are hitting, and seeing if there’s some way I can gather the right data to gain insights, even if they seem hopelessly qualitative at first glance.
User education is painful but powerful
HubSpot focuses on the sort of people who used to buy ads in the Yellow Pages to promote their businesses. These people know they now need to use the Internet to reach customers, but they aren’t sure how. To succeed, HubSpot has to help those people build useful websites and channels. Templates and other automated tools help, but a lot still comes down to people creating the right content for their own businesses and responding appropriately when customers get in touch through Twitter, Facebook or email. The only way to achieve that is to teach people how to do it, and so a lot of the company’s resources are put into education.
On a simple level, tools like HubSpot’s graders offer simple suggestions for improving websites and other content. Users of the service are sent regular emails that remind them of steps and actions they need to take, such as updating their blogs. HubSpot hosts a popular video cast that covers all sorts of tips and horror stories from the last week in social media. All of these efforts really seem to help the company, judging from how enthusiastically users respond to all the material. On a deeper level, it also seems to help build a long-term relationship between the company and its customers, driving real loyalty.
One of the unwritten rules of the consumer technology world is that anything that requires educating users is a losing proposition. Anybody who has looked at their customer acquisition funnel knows how even minor usability problems can drive away vast swaths of people. What’s different about HubSpot is that their customers are a lot more motivated than your average consumer on the web. They’re using the service in the hope of actually making more money, so they’re willing to invest some time. It left me wondering if I should spend more time creating training material for my own projects, rather than always prioritizing interface work to make them easier to use. The people who use them to create content are already investing their own time, so is that perhaps another situation where education would pay off?
Hubspot is a smart, practical company that’s very focused on using the data they’re gathering to understand what their customers really need. Maybe that’s precisely because the team isn’t in the Valley to be distracted by every shiny new idea? No matter what the cause, I’m grateful that they spent the time to show me what they’d learned, and I’m looking forward to applying these ideas to my own work.