In this interview I will share the interview I had with Snowplow Analytics co-founder Yali Sassoon. We discussed the changing digital analytics solutions landscape and how Snowplow Analytics fits in the landscape as a disruptive digital analytics solution.
Snowplow Analytics is a datawarehousing platform for web and mobile (and any other kind of digital event) data.
You integrate Snowplow tags or a mobile or server side SDK in your website, mobile app or server side application, just as if you were instrumenting a standard analytics package. Instead of providing you with a UI to access your data, we deliver your data to you, in a very granular, structured format (one line of data per event) in your own data warehouse. You can then do whatever you want with it.
What are the major challenges Snowplow analytics is solving for digital analysts that other web analytics solutions are not?
Traditional web analytics tools are built to deliver a narrow set of reports, with a particular focus on customer acquisition and conversion. This is a real shame, because digital event data is incredibly rich data – it tells you how real people have interacted with your websites and applications, in an enormous amount of detail. There’s a huge amount you can do with that data – to give just some examples:
- Unpick to what extent behaviour is driven by who a user is, what he / she is trying to accomplish, and how you’ve architected your website or application
- Segment your audience based on who they are and what they’re trying to accomplish, then measure how successful you are at meeting each segment’s needs
- Compare how your products (if you’re a retailer), articles (if you’re a newspaper site), media (if you’re a video site) perform in terms of attracting and engaging users
- Spot the events on a user’s journey that are predictive of higher value events further on in that users journey, and explore different ways you can engage with your users differently to grow their value
Snowplow frees data analysts to start performing these higher value analyses, by giving them the data in a format that makes it possible to:
- Join with any other data set (e.g. CRM, marketing, merchandising, CMS, financial)
- Plug in any analysis tool
What recommendation can you give web analysts when it comes to web analytics tool selection? What should they be particularly focused on or give attention to?
I think there’s a journey that companies need to go through as they mature in their use of data. Most companies start by performing basic analyses with free tools – namely Google Analytics, and this will get them so far. At a certain point, some companies (but not all) will grow frustrated because:
- Their web analytics data is siloed – so very hard / impossible to integrate with their other data sets
- Their access to the data is mediated via Google’s APIs, and they cannot get at the underlying event-level data, only aggregate slices of that data
At this stage the company will have three choices – they can:
- Migrate to Google Analytics Premium
- Buy the Adobe Marketing Cloud
- Setup Snowplow, generally alongside a Business Intelligence tool. (Which provides the front end for most users querying the data)
I would always go with (3), because:
- You own your own data – and I believe that the companies that win in tomorrow’s market will be the ones that treat their data and the insight built on it as a core asset
- Your digital data ends up warehoused alongside all your other data sets – providing you with the best possible basis for making data-driven decisions
- It’s the most cost effective solution
I am, obviously, rather biased :-). But I co-founded Snowplow because I believe in the above, it’s not that I believe in the above because I work for Snowplow.
What is a recent development or trend in digital analytics that got you really excited about?
Real-time data processing is incredibly exciting. Not from a reporting perspective, because as soon as you have an analyst consuming reports, the analyst will be the bottleneck on data being actioned, rather than the speed with which the data is delivered to the analyst. Rather, real-time data processing is exciting because it makes lots of data-driven products possible: for example, product recommendation engines, personalization engines, marketing optimization engines.
At Snowplow, most of our engineering effort is now being put towards working on our real-time pipeline: so that as well as loading data into a datawarehouse, we push data, in real-time, into a Unified Log (on Amazon Kinesis) that means our users can write data-driven applications to read off the log and take actions based on it. That means that the role of digital analyst will change from providing insights to drive business decisions to building, testing and iterating data-driven products.
Do you think digital analysts are open to accept disruptive change of the digital analytics solution space? Why do you believe this?
There are all kinds of digital analysts – some of whom are excited by disruptive change, others of whom are unsettled by change.
I think the bigger barriers to disruptive change are at the company level, rather than the analyst level, however. Integrating a new analytics solution is risky:
- Implementations go wrong, especially if there aren’t the right resources behind them
- Having another source of truth creates discrepancies that have to be explained
- The data might not tell you what you want to hear
For that reason, we see most new Snowplow users adopt our technology to solve a specific problem that they can’t solve with their current provider, rather than to replace their current setup. Then, once they get more familiar with our data and become more expert at using it to answer business questions, they start to use Snowplow at the expense of their old tool.
If you look towards the future, what will be the next steps for Snowplow analytics?
Our immediate focus is on making our real-time flow production ready.
There is then several years worth of working with our clients to build a toolset so that they can put together the algorithms and data-driven tools they need to power their businesses.
In parallel, we want to integrate with as many different service providers in digital. This is a two way process: we want to integrate, for example, with third party email marketing programs so that data on what email is sent to who can be easily recorded into Snowplow. (We already have an integration with Mailchimp and Mandrill that does this via webhooks.) At the same time, we need to make it possible for our users to build data-driven applications that target individual users based on the pattern of events recorded in Snowplow and trigger an email to respond to that pattern.
Snowplow Analytics is coming to the Netherlands
On May 13 Snowplow Analytics will host the first Snowplow Analytics meetup in beautiful Amsterdam. This will be a great opportunity to meet with Snowplow Analytics and learn more about the solution. There will also be presentations by Snowplow clients to share how they are using Snowplow Analytics data in their data stack for optimization, analysis and reporting. Featured clients are De Bijenkorf and Blue Mango.
If you would like to join this free meetup, please sign up here. Don’t wait! Limited seats available.