Wow! What a great conference at Structure Data this week in New York.
Two things I love to do at technology conferences are: 1) walk around the exhibition hall observing how vendors present themselves; and 2) talk to the passersby at our own table. This helps me get a better understanding of the industry as well and helps me understand where Metamarkets fits in.
Tools vs Solutions
First, here are some general observations about the technology landscape for Big Data:
- Before joining Metamarkets, I spent a lot of time talking to numerous companies and individuals about the industry. There was one particular observation from a VC friend that struck me this week. Typically, new companies fall into one of two buckets: (i) addressing an existing category by being better, faster, stronger; or (ii) solving a new problem that, if successful, becomes a new category over time. There are a lot of companies focused on making Hadoop better, faster, stronger.
- A lot of vendors present were offering tools instead of solutions. Splunk and a few others (including us) were the exceptions.
To that end, here’s a table that I’ve developed that describes the differences between Tools vs Solutions.
The Pain of Data Bureaucracy
Many who stopped by our table had heard of Metamarkets but didn’t know what we did. Let me answer here by first describing a hypothetical chain of events an online company may experience:
- Acme.com has dashboards and charts (built with Excel, Tableau, or Spotfire) to their track key performance indicators (revenue, impressions, check-ins, clicks, likes, or other transactions, depending on the industry). These metrics are updated weekly.
- One week, there is a huge drop in one of Acme.com’s key metrics. The management team wants to know why.
- Behind the scenes, this question results in a flurry of requests to the analytics team. They run a series of SQL queries, Hadoop jobs, and additional reports.
- Several iterations may be required based on miscommunication, misunderstanding of the original question, and subsequent follow-on questions.
- This entire cycle may take days, only to be repeated the next time there is another spike or drop in a key metric.
- Frustration ensues because the business cannot get key answers in a timely fashion. The analytics team is overwhelmed, and worse, underutilized: they are servicing these requests instead of conducting deeper analysis.
Sound familiar? Maybe the company down the street?
This is the pain that Metamarkets addresses: enabling end users to monitor key metrics and answer their own questions. This results in better, faster decisions and frees up the analytics team to conduct the rich analysis they were hired to perform.
What Metamarkets Does
While we have great technology, our focus is on delivering a solution to this painful data bureaucracy. Specifically, Metamarkets excels at making sense of time-dependent, event-based data. Some customer scenarios include:
- Segmenting user actions (check-ins, likes, favorites, purchases, tweets, etc.) according to demographic information such as geography, age group, venue location, and gender among others.
- Comparing overall user performance across previous periods and user segments.
- Improving usage by measuring the impact of new features on user actions.
- Improving pricing decisions by comparing historical supply and demand with current inventory levels.
- Identifying where advertisers are being outspent by their competition.
- Demonstrating campaign effectiveness based on comparables and prior results.
- Getting a unified view across ad server, revenue, demographic, and audience targeting data.
- Quickly seeing how revenue and other key metrics are tracking compared to plan as well as previous periods.
- Generating new revenue streams by offering advertisers additional services and audience targeting opportunities.
The other question we frequently got at the conference was how Metamarkets compared to various front-end visualization tools or back-end columnar data stores. The detailed answer will have to wait for another blog post, but the short answer is: they offer tools, whereas we provide a solution.
Are there other time-dependent, event-based, data-related questions you think Metamarkets would be well-suited for? We would love to hear from you.