Becoming a data-driven company

-


How does a company become data-driven? Here's what I've learned at Zervant.

Radiate information


If you want to turn the culture more data-driven, radiate information. You'll never know what will stick. Radiating information also has the side-effect of evangelizing the work of analytics: people will remember you exist and they will take analytics into consideration more easily.
Oh right, when we're building this we need to talk to analytics about what to measure! 
Write an internal blog, give short presentations about an information nugget, etc. Frequency is more important than the perfect quality of content.

Teach


People love to learn new things. They will first tell you they're in no way analytical-minded and far from technical, but they will eagerly learn. De-mystifying analytics will help everyone in your organization to speak the same language and be on the same page about what analytics can do for them.

As I've held workshops on analytics, I've seen people light up and apply their newly discovered analytical mind in a sophisticated manner on their own fields including HR, branding, marketing and customer support. It doesn't take much, but it gives back a lot.

Be exemplary in prioritizing projects by their business potential


Who doesn't want to develop a fancy neural network that predicts accurately the next user to convert? But if you ask yourself what you can really do with that information and you struggle to answer, it's a sign that you should perhaps reconsider your priorities.

Everyone in the organization will have opinions on what the most important thing is. Amazingly, the most important thing tends to depend on the person and astonishingly often is related to their own field. This applies to us in data science too. Break free from that, be objective, realistic and show the right kind of example to the rest of the organization. If you have good, concrete arguments of a projects potential, it will be hard to oppose it with only opinions.

Create a data strategy: start from what your business needs, identify the points that have the largest impact, but equally importantly brainstorm new opportunities for use of data. Make the business case clear and be honest about it.



Experiment lightweight


It doesn't matter that you don't have a fancy A/B testing suite, you can still do lightweight experimentation. Are you planning to add a new feature? What if before implementing any of it you just create a mock up with a button to check who clicks it, who's interested? If you do this, don't be a douche. Do it with respect to your users.

The point is not to get it right the first time, but rather to show how easy it is to test something out. This will help people understand the power of lightweight experimentation and they will have lots of new ideas to try out.

Celebrate even the small wins


Everyone loves a good story, even if the story is about how the definition of your active user got improved - as long as you make it relevant to your audience. Celebrating even the small wins is not just good for morale, but it also highlights the impact of this work. In order to continue enjoying the trust of your top management, this is critical.

You won't make it alone


Data-driven doesn't mean you shove data down everyone's throats. Rather, try to find approaches that involves others. Analytics is woven into everything a company does. When it is done as an afterthought, it becomes an isolated activity - something to sprinkle on top by the analytics team before things go live. Preventing this from happening unfortunately means having good personal relationships with the key influencers in your organization. So work on your social networking skills.

Comments

Post a Comment

Popular posts from this blog

How to access AWS S3 with pyspark locally using AWS profiles tutorial

-
At Zervant , we currently use databricks for our ETL processes, and it's quite great. However, there's been some difficulty in setting up scripts that work both locally and on the databricks cloud. Specifically, databricks uses their own prorpietary libraries to connect to AWS S3 based on AWS hadoop 2.7. That version does not support accessing using AWS profiles. Internally, we use SSO to create temporary credentials for an AWS profile that then assumes a role. Therefore, reading the ACCESS_ID and ACCESS_SECRET from the .credentials file is something we don't want to do. In order to accomplish this, we need to set two hadoop configurations to the Spark Context fs.s3a.aws.credentials.provider com.amazonaws.auth.profile.ProfileCredentialsProvider This is done by running this line of code: sc._jsc.hadoopConfiguration().set("fs.s3a.aws.credentials.provider", "com.amazonaws.auth.profile.ProfileCredentialsProvider") Note! You need to set your environment var...
Read more

Snowflake UPSERT operation (aka MERGE)

-
You want to insert data to a table, but if a corresponding row already exists (by some rule, e.g. unique key) you want to update that instead of adding a new row, keeping the dataset's unique requirements intact. That's an "UPDATE AND INSERT" operation, or UPSERT. Some SQL languages have native support for it.  PostgreSQL has UPSERT as native. Also MySQL supports the operation with INSERT and ON DUPLICATE KEY UPDATE. How do you do UPSERT on Snowflake? Here's how: Snowflake UPSERT i.e. MERGE operation Snowflake's UPSERT is called MERGE and it works just as conveniently. It just has a different name. Here's the simple usage: MERGE INTO workspace.destination_table d USING workspace.source_table s ON d.id = s.id AND d.val1 = s.val1 WHEN MATCHED THEN update SET d.val2 = s.val2, d.val3 = s.val3 WHEN NOT MATCHED THEN INSERT (id, val1, val2, val3) VALUES (s.id, s.val1, s.val2, s.val3); Here the destination_table and source_table are of similar form,...
Read more