Faced with a host of challenges from a growing amount of data, HomeToGo adopted an analytics platform from Sisu to turn its data from an obstacle into an asset.
Berlin-based HomeToGo is a marketplace for vacation rentals that operates 23 travel websites and applications including Tripping.com and Casamundo, with a marketplace that includes listings from partners like Booking.com, TripAdvisor and Vrbo.
By 2021, seven years after its launch, HomeToGo was collecting more data than it could handle. Data sources include events from its many websites, like searches and purchases, along with external marketing data provided by third parties.
But rather than serve as a starting point for analysis and insight, the data was becoming paralyzing.
According to Athene Cook, HomeToGo's senior data analyst, with so much data on hand the organization's business users didn't know where to start to gain insight for marketing campaigns and performance analysis.
They could come to her team for assistance, but that created a bottleneck that delayed any analysis for days, until the data team could get around to helping with a particular request.
In addition, Cook noted that the volume of data made it difficult to surface the most relevant data and made it challenging to analyze key metrics like property and market performance to understand why a particular property or market might be in high -- or low -- demand.
"There could be a million different factors that could be causing something," Cook said during a webinar hosted by Sisu on Jan. 27. "It's really tricky, as a human, to know which of them is most likely to be most important and where to even begin slicing and dicing. … Best case, going from exploration to discovery is going to be a couple of hours, but more realistically we're thinking about days."
Athene CookSenior data analyst, HomeToGo
That lag time -- which doesn't even factor in the delay between a query request and when the data team could begin its work -- results in less productivity and, ultimately, lower revenue, Cook added.
Rather than add more data analysts, HomeToGo determined it needed a self-service analytics tool that would reduce the stress on the data team by enabling business users to surface relevant data and use that data to take action.
"We didn't want to throw bodies at the problem," Cook said. "That doesn't scale, and it's not a very good use of people's expertise. We needed something that, first and foremost, was going to reduce time spent on manual slicing and dicing. Second, we wanted to make sure we were promoting a self-service analytics mindset throughout the organization."
A solution in Sisu
In addition to increased speed-to-insight and self-service capabilities, HomeToGo's third criterion when seeking an analytics platform was avoiding vendor lock-in. The company wanted an analytics platform that would enable it to remain agile, not forcing it to connect to one cloud platform versus another.
"We have the philosophy that we want to find a tool that's best suited to getting a certain job done rather than trying to find a Swiss-army-knife solution that does everything," said Cook, who noted that HomeToGo recently switched cloud data warehouses from Amazon Redshift to Snowflake.
And that means implementing tools that work agnostically with others, she continued.
After trying out Sisu in the summer of 2021, HomeToGo determined that the startup's analytics platform met its criteria and was the right fit. Sisu's platform uses augmented intelligence and machine learning to monitor data sets, automatically alert users to any changes in those data sets, and then explain why those changed occurred. Since emerging from stealth in 2019, Sisu has added analytics capabilities that enable users to ask questions and understand what to do next.
Time-to-value was a key consideration for HomeToGo, so it tested Sisu's platform against known outcomes by applying Sisu to query, analyze, explain and action data that had already been queried, analyzed, explained and acted on by HomeToGo's data team.
"Needless to say, it passed the test," Cook said.
According to Cook, what had previously taken days and weeks was accomplished in less than a day using Sisu. And that, she pointed out, was HomeToGo's first time using an unfamiliar platform.
Even after Sisu was selected, challenges remained, Cook noted.
HomeToGo had to onboard Sisu, which meant training its business users to use the analytics platform and become more self-sufficient. It did so initially through a company-wide training program, and then further training of a group of "super users" within different departments to make them the go-to people for business users when questions arise.
"If there are other people on their teams who are new to Sisu or have questions, we in analytics are no longer serving as the bottleneck," Cook said. "They can approach these super users in their domain and get their domain-specific help."
In addition, super users now essentially serve as evangelists for analytics within HomeToGo, she added.
"They've become Sisu culture ambassadors across HomeToGo," Cook said. "We in data are always talking about how great these tools are and people should use them, but it makes all the difference in [adoption] when you're also hearing it from a business user."
Sisu use cases grow
Since it's only been a few months since HomeToGo deployed Sisu, the analytics platform has primarily been used for marketing purposes, according to Cook.
In addition, the company's product team has employed Sisu to do A/B testing to help determine which version of a product drives more business.
Going forward, however, HomeToGo envisions more analytics use cases for Sisu.
According to Cook, the company plans to take advantage of Sisu's application programming interface (API), specifically to gain more control over scheduled analyses that run daily. On rare occasions, the analyses are scheduled to run before the underlying data is available. Via the API, HomeToGo would be able to instruct Sisu to delay the analyses and therefore avoid any false alarms.
In addition, HomeToGo plans to use Sisu to enhance its reporting capabilities, complementing reports with daily emails that, for example, list the top 20 drivers of revenue.
"As we like to say, that's going from the 'what' to the 'why'," Cook said.
While Sisu specializes in deploying machine learning algorithms to explain why something occurs within an organization's data, it is expanding its capabilities. Meanwhile, other vendors have begun adding capabilities that attempt to explain why something occurs. Sisu's competitors, therefore, now include some of the more established analytics vendors such as Looker and Tableau.
Sisu *pricing is not publicly available. Competitor Tableau offers multiple pricing plans ranging from $12 per user, per month to $70 per user, per month. Looker does not publish its pricing information.
*Incorrect pricing was provided when this story was first published. We regret the error.