5 New Year’s resolutions for analytics
As we edge closer to 2017, columnist Nick Iyengar discusses key goals you should consider for your analytics program in the new year.
The holiday season is upon us, and with that comes the usual cycle of overindulging at the dinner table, followed by New Year’s resolutions to get back in shape. And while I, for one, plan to do both of those things, I think it’s also a good time to start considering some New Year’s resolutions for your analytics program.
Based on my experience consulting with clients in 2016, I’ve jotted down a handful of ideas for resolutions I think you’ll find useful for 2017. Fortunately, all of these are a good bit more realistic than getting back in shape after a month full of pie and eggnog!
1. Review your goals and KPIs
Yes, we’re starting with the basics. One of the most fundamental reasons analytics loses its ability to drive performance improvement is that measurement can become detached from decision-making. So, get your stakeholders in a room (physical or virtual will do), and ask them point-blank: Are the metrics we’re giving you relevant? Do they help you make the key decisions you need to make?
Candid feedback on this most fundamental building block of analytics can help you identify missing KPIs, existing KPIs that don’t add any value in supporting decisions, and those that stakeholders aren’t confident in, don’t trust and so forth.
Based on a review session like this, it’s likely that you’ll need to update your dashboards, adjust your reporting templates and make other changes. But doing that work to spruce up your analytics “product” within your organization is well worth it to avoid having your reports sit on shelves, unread.
2. Replace manual reporting with automated dashboards
Once you’ve confirmed that the metrics you’re using to measure performance are still relevant to the decision-makers using the data, you can feel confident that you’re on the right track. But if you spend a lot of time pulling data from your data sources, compiling and wrangling it in Excel, and building charts and PowerPoint decks, knowing that all of that effort is producing relevant data may be of little solace.
To break the cycle of labor-intensive manual reporting, consider how automated, “always-on” dashboards can help you.
The good news is that there is a myriad of dashboarding tools that can help you, with lots of features and tools to choose from. You can pick out a basic, functional “Toyota” as your dashboard tool, or go with an exciting, flashy “Ferrari,” depending on your needs.
But the basic idea is this: Rather than going through the same laborious reporting process every day/week/month/quarter/year, instead build out a dashboard (or set of dashboards) that presents the same data, that is accessible to all necessary stakeholders, and that is always on, so that people can review it at their convenience.
What will you do with all of the time you’ve just freed up? Well, how about doing deeper analysis — not just reporting out on how many page views or transactions or upsells or MQLs (marketing-qualified leads) were generated in the last period, but digging into why that happened and how to improve in the future.
If done well, your always-on dashboards will free up a lot of time for analysis and also generate demand for that deeper analysis from your stakeholders. That’s a virtuous circle!
3. Audit your data to uncover issues
You’ve probably heard the saying that “an ounce of prevention is worth a pound of cure.” (Incidentally, is there a metric equivalent for this phrase?) And that’s certainly true when it comes to ensuring the integrity of the data in your marketing stack.
It’s much better to proactively scrub your web analytics data, your CRM data and so on, and discover that everything is working flawlessly, than it does to be blissfully ignorant of data quality issues lurking in your systems.
Auditing your data is relatively quick and easy and promotes broad trust across the organization that the data can be used to make important decisions. To avoid an audit that ends up as a rambling, ill-defined “excursion” through your data, consider putting together a checklist before you begin. A checklist should give you specific issues that you’ll be checking on, and it should define criteria for whether each issue “passes” or “fails” its check.
To use Google Analytics data as one example, you might consider auditing your event tracking data. Your criteria for “pass” might include, for instance, “no unexplained weekly swings of +/- 90% in event volume for any given Event Category.” Note the importance of the word unexplained. There may be legitimate reasons why any given metric might make a sudden, major move. As long as there’s an explanation, you’re OK. If there isn’t, it’s time to dig deeper and figure out what happened.
And as simple as that sounds, root cause analysis is key to any worthwhile audit. You don’t want to simply bring issues to the attention of data consumers. Instead, it’s crucial to get to the root of the issue so that it can be resolved, so that similar issues don’t recur in the future.
4. Data integrations: 1+1=3
If you’re in the enviable, and dare I say unrealistic, position of having no data quality issues that need to be hammered out, consider moving on to a more exciting question: How can you raise the value of the data contained within individual silos by integrating your datasets?
This is the “1+1=3” scenario: By finding ways to merge distinct datasets, you may be able to take action on your data in ways that were never possible before.
Say you’ve got your customer data in a CRM. You’re tracking things like Industry, Company Size, Job Title, and the like, if you’re a B2B organization, or things like lifetime value and number of purchases, if you sell to consumers. Then, you’ve got all your web activity data in a tool like Google Analytics or Adobe Analytics. Without a connection between these two silos, all of your web data remains anonymous — you don’t know if those browsing the site are high-value leads or people who could never become customers at all.
Integrating these two data sources enables much richer segmentation. Suddenly, you’re able to look at web data through the lenses of interesting segments, like “Leads with a High Lead Score” or “Customers with over $500 in Lifetime Spend.” That more advanced segmentation is key, because it leads to more specific insights, as well as to tighter targeting — for instance, in remarketing campaigns.
Of course, tighter targeting leads to more personalized communication and, at least in theory, stronger ROI.
So if you’re feeling good about yourself because you couldn’t find any wonky data in Salesforce, or because your data in Adobe Analytics all checks out, don’t get too comfy. You can raise the value of your organization’s data by tackling the challenge of data integration.
5. Get your staff trained on your key tools
Everyone likes to talk about new tools and technology. Nobody likes to talk about training. Training is the “vegetables” of the Analytics Food Pyramid (patent pending). You know it’s good for you, and it’s really cheap, but are you getting enough?
Here’s the thing about training: I routinely work with organizations that aren’t getting the maximum value out of the tools for which they are paying. Often this is due to bandwidth limitations (“We just don’t have enough people to fully use this tool.”) or simply knowledge gaps (“Nobody here is really sure how to get the most out of this tool.”).
Here’s the good news: The end of a year, and the start of a new year, is a great time to carve out some space for your staff to go get that training they need to ensure you’re getting serious ROI from the technology you’re already paying for. On top of that, training is a great way for people to check off those “personal development” boxes on a performance review.
As a consultant, it’s very common to get pulled into a vendor evaluation for a new piece of technology. I often find that organizations put vendors through a rigorous evaluation to ensure that the technology truly fills an identified gap, and that’s great. But it’s much rarer to find an organization examining itself critically to determine whether it can staff a new technology with the people and training needed to extract the potential ROI that tech vendor will be pitching.
Don’t let your technology collect dust due to a lack of use. Don’t buy a tool and then use it for 10 percent of what it can do. Get your teams trained on the tools already at their disposal so that you’re confident you’re getting maximum ROI from those tools.
So, as you ponder a new home gym when you start thinking about New Year’s resolutions, spare a thought for some of the analytics resolutions I’ve offered here. They might not burn off Mom’s pumpkin pie, but they’ll help you make sure 2017 is your best year yet in analytics.
Opinions expressed in this article are those of the guest author and not necessarily MarTech. Staff authors are listed here.