Analyst skills are now table stakes: how to be dangerous with data
Strong digital analyst skills separate the good from the great knowledge economy workers ...some ideas on upping your game
As digital analyst is a fast growing role at startups and large companies, one with a ton of opportunity for new and experienced analysts alike, today I thought I’d share some ideas on how you can become a better analyst. Also at this point, defensible analyst work should really be a skill every knowledge worker and business executive possesses, not just your data science and marketing teams.
Improving our digital analysis skills goes beyond broadening career options and helping us be better at our craft. It should actually improve all areas of life as a byproduct of nurturing critical thinking skills. Some recommendations follow I think are valuable for everyone (and that I personally subscribe to myself).
Find a passion outside work which involves developing hypotheses
The scientific method, as we all know, is a body of techniques for investigating phenomena, acquiring new knowledge, or correcting and integrating previous knowledge.
You’re already applying this to your business and analytics practice by putting it to work for testing and optimization efforts (for example, having a hypothesis that a new landing page with less clutter or a new ad campaign with a snappy tagline will convert better, which you then test).
But beyond work, you should also, in free time, be involved in something which flexes your prediction muscle.
Whether this manifests as fantasy sports, investing in startups/stocks or some other activity which involves future predictions (and cool datasets!), this can be a fun and rewarding way to sharpen your mind and will help you see analysis problems in a new light.
Have sandbox projects to test new tools
If you are truly serious about improving your skills, doing analyst work in your live business environment isn’t enough.
The reason being you can’t test and tinker with any tool without permission or change settings at whim, you likely have compliance and managers to work through. You’re also using live data being captured that if you’re new to experimenting with code/config you don’t want to risk breaking anything.
But a sandbox project such as your own site, app or side business provides a place you can test, tinker and experiment in a no-stress setting.
Bonus: my former team at Google previously launched an Analytics Demo Account you can do anything you’d like with to get comfortable without fear for this very purpose along with an accompanying online analytics course which is free to take, on-demand.
Learn to fill in the missing pieces, be comfortable working with imperfect data/information
100% perfect data is really only possible in a controlled lab setting with expensive and fine-tuned equipment. While, of course, we should ensure our analytics implementation is setup correctly to keep our own data as clean as possible, we must also get comfortable working with a “good enough” information, especially as we rely on more tools from platforms and aggregate/directional datasets such as Google Trends.
Being able to (intelligently and as unbiased as possible) fill in the missing pieces is necessary in order to be agile in how we work and keep projects moving forward. A great analyst will work out the way to fill in blanks and still make effective projections (while of course providing a rationale/caveats where needed and always adding annotations).
You want to get confident enough and have your team be confident enough to make recommendations and create analysis’ based off “minimum viable data.”
Learn when and how to run A/B and multivariate tests/content experiments
Your landing page is performing well, but you can’t help but think it could do better. You could replace it with a new design, but how do you know if it’s truly a better experience? Worse, what if it performs poorly? You’ll lose revenue and conversions while you’re busy running the numbers. There’s a better way: use a system to test your content.
Google Optimize, a free testing tool, lets you compare how different pages perform using a random sample set, define what percentage of visitors are included in an experiment, test different objectives and get automated updates on how experiments are doing. And the best part about is it works to get a statistically relevant sample set as soon as possible, to minimize the number of users who see a poorer performing page — all automatically.
Regardless of which tools you use, the point of is that you can leave the hard work of testing and experimentation to software and free up your time for more strategic and creative work that only humans can do. I highly recommend creating a testing/optimization roadmap to run tests: my former colleague Krista Seiden generously published a free basic one you can make a copy of for your business here.
Learn when to recommend actively using data in creative ways: ex, using a real-time API
Most popular brands always have visitors on their site. While there, these visitors are viewing content, sharing things they find interesting with their social networks, making purchases and more. But are you using these actions to improve the experience of other users?
Looking at this data in real-time and taking action on it is not always feasible. But, using a Real-Time API to dynamically improve your site content is. Using an API from your analytics provider, you can make queries about your real-time data and use that information to create a more compelling experience.
One of the most basic but practical use cases is to manage the content on your webpage. For example, you can query the API for the top visited URLs to construct a “top trending content widget” and display the number of active readers. This can help create a greater sense of urgency to view and share ‘hot’ content.
Additionally, this metric can be shown on different conversion pages of a website to impart a sense of urgency and demonstrate demand for a given product. Twiddy, a family-owned vacation rental company, with the help of their consultant Joe Akinc, has been testing this. Not only did their revenue increase 18.6 percent, but the average order value increased 11.9 percent and the conversion rate increased 7.9 percent. See the Twiddy case study for the full story.
Learn which data firehoses you have access to in your own organization or can pay for, and how they might be applied to improve your company’s processes or products.
Know how to use your site analytics to power critical tools such as remarketing
Ever visited a website intending to make a purchase but got side-tracked halfway through with another task? Of course you have. As the above image shows, it’s happened to nearly everyone. Wouldn’t it be nice if you could give a friendly reminder to a user to help them complete their purchase? This would be a herculean task manually, but programmatically it’s much easier.
Remarketing with Analytics lets you follow up with people who have already visited your website, and deliver ad content specifically targeted to the interests they expressed during those previous visits.
Nearly all sophisticated analysts and marketers use this powerful tool, just be careful to not abuse remarketing in such a way as to frustrate or saturate your users. Great analysts can help you get the frequency correct in a way that is not intrusive.
Live and breathe your company and sector metrics (beyond what you’re accountable for)
Being a great analyst isn’t about just running reports and delivering insights that are your remit.
Rather, the best analysts have their finger on the pulse of the bigger pictures and are deftly able to put their own work into context with the larger organization and their sector as a whole.
The analysts I talk to that leave a lasting impression are the ones who can speak articulately about various areas of the organization and how they make impact across teams and functions, and deeply understand the business they’re in.
Be an active member of the industry, network and collaborate with peers
I’m personally a big believer in educating others about digital marketing and since starting my career in the corporate world in 2005 I’ve spent time both at and outside of work learning from peers as well as helping others learn our craft.
Our industry is tight knit and so being an active participant who helps others is of great benefit (not to mention fulfilling).
For you, whether this takes the form of speaking/attending events, starting your own local analytics meetup, publishing case studies on a blog/Substack/industry trade or even making friends with other analysts on Twitter to talk shop, this is a valuable use of time.
Bonus: if you’re looking for detailed guides, following are 4 analytics books I recommend
1. Successful Analytics — gain business insights by better managing analytics
Change the way your organization perceives Analytics — this book by analytics expert Brian Clifton puts you in control of your data showing you how to use Analytics strategically in ways that connect directly with your company’s bottom line. Aimed at data managers and senior analysts, it provides detailed guidance on the entire process of building a data-driven environment using analytics and then applying it to produce real-world insights.
2. Visualize This — the flowing data guide to design visualization and statistics
There are lots of books on visualization that describe best practices and design concepts, but what do you do when it comes time for you to actually make something? If you don’t know how to use the software in front of you, the abstract isn’t all that useful. And with growing amounts of data, it’s becoming more important to be able to make sense of and communicate with it all. In Visualize This, popular blogger Nathan Yau teaches you how to create graphics that tell stories with real data, and you’ll have fun in the process. Learn to make statistical graphics in R, design in Illustrator, and create interactive graphics in JavaScript and Flash & Actionscript.
3. Web Analytics 2.0 — the art of online accountability and science of customer centricity
Adeptly address today’s business challenges with what many still consider bible on web analytics book from web analytics thought leader and Googler Avinash Kaushik. Web Analytics 2.0 presents a new framework that will permanently change how you think about analytics. It provides specific recommendations for creating an actionable strategy, applying analytical techniques correctly, solving challenges such as measuring social media and multichannel campaigns, achieving optimal success by leveraging experimentation, and employing tactics for truly listening to your customers.
4. Building a Digital Analytics Organization — Create Value by integrating analytical processes, technology, and people into business operations
This guide written by Judah Phillips covers all you need to know to build a well-resourced digital analytics team, and then back it with cross-functional support and alignment from IT, marketing, finance, executives team, and beyond…while successfully applying analytics across the business. You will learn what it means to be “doing analytics”: creating analytical processes and managing teams; collecting and governing data; analyzing paid, owned, and earned media; performing competitive and qualitative analyses; testing and optimization; targeting and automating; integrating digital data; using predictive modeling and other data sciences; and much more.
If you ever have questions on broadening your digital analyst skills, you can always follow me on Twitter and ask, am always happy to try and help.