Time to Switch Analytics Tools?
Migrating analytics systems can be a serious endeavor. Organizations will switch over to Adobe Analytics or Google Analytics 360 for a variety of reasons including cost savings, tooling capabilities, staff expertise and compatibility with marketing tech.
If you have been a faithful fan of one web analytics tool, getting set up on a new one takes time and expense. Learning a whole new set of processing rules, admin settings and naming conventions can be a steep curve. And, all of that comes after the project management and costs typically needed for a seamless transition. We at Empirical Path have led several of these migrations, typically taking clients from Adobe Analytics to Google Analytics. Below are several of our favorite recommendations for planning a smooth and successful transition.
Why Change Analytics Platforms?
Before you begin, ask yourself and your stakeholders why — really why — you are making the jump to Adobe, Analytics 360, or some other tool.
Financial constraints are always a factor, but with that aside make sure to do your due diligence on whether your newly selected tool has all of the sought-after features to answer your critical business questions.
Marketing stack or digital ecosystem is often cited as a critical consideration. Google Analytics, for instance, offers seamless integrations within Google’s stack of tools, including AdWords, Optimize, Data Studio, Search Console, and, for Analytics 360 accounts, BigQuery and DoubleClick. Adobe offers similar integrations within its Marketing Cloud products. While not impossible to cross-fertilize between some marketing platforms, try to check the box on as many feature sets and capabilities as desired and think through the cost-benefit of foregoing others.
Document, Document, Document
While hasty documentation can be fatal, spending extra time to get it right at this stage delivers extra benefits in the long run. When leaving Adobe, for example, make sure you have an updated Solutions Design Requirements (SDR) or similar document detailing what data you currently collect. Examine this document thoroughly and decide what eVars, props, events need to be captured in Google Analytics. This is a good time to rethink what is essential to answering stakeholder’s questions by setting KPIs and cutting out fat to reduce development and QA efforts.
Make sure you are familiar with the limits of each platform. If you are cutting over to Google Analytics Standard, remember you’ll be limited by the number of custom dimensions (20 for Standard Edition, 200 dimensions in Analytics 360), whereas Adobe offers a higher capacity to capture custom traffic data (eVars).
Standardize First with Tag Management and a Data Layer
Technically, a tag management system (TMS) is optional, but not really. If you aren’t already using a TMS combined with a well-defined data layer to feed your preferred analytics tool, start now. The many benefits of a data layer to any serious analytics setup are well documented elsewhere, but the importance of data standardization to a web analytics migration deserves special mention here.
In addition to sending data to a preferred web analytics tool, most organizations also supply web interaction data to other platforms like ESPs, Display & Retargeting vendors and customer systems. Standardizing how you collect which information via the data layer eases future transitions and integrations. If you are lucky enough to have a solid data layer in place already, then transitioning between tools becomes a simple matter of editing existing rules within your TMS to supply multiple vendors.
Run Both for Quality Assurance and Happy Stakeholders
If you have the ability, consider overlapping Adobe and Google during a settling-in period for quality assurance purposes. There will always be stakeholders within your organization upset or anxious about how changing platforms will impact their work. Increase comfort and confidence while setting expectations by running your tools simultaneously to showcase that the new system can capture similar data sets, albeit presented differently. Try to guardrail the comparisons to key metrics and dimensions to avoid heading down the rabbit hole of figuring out each nuanced difference.
Handling historical data from a legacy system can get messy, so start by answering these important questions:
- What data should I store?
For instance, if starting from an existing Adobe setup, ask yourself if it’s worthwhile to collect all of your eVar and event data and with which dimensions — marketing channels? VisitorsIDs? Geography? To reduce complexity, think about how and what you report on today and only pull necessary data. On the flip side, if you have development resources and cost isn’t the main concern you could send Data Warehouse exports to an FTP and process into a database. Alternately, you could stream Data Warehouse into a database using its API. For session-level data exports out of Adobe Analytics, another helpful option is the Data Feed feature (formally Clickstream data feed).
- How much should I store?
Will you need data with daily/weekly/monthly granularity for the past one, two, or three years? Again, think about the current need; if you are only reporting year-over-year stats, will storing data from 3-5 years ago be worth the effort? Is the data from that historical period relevant to your current digital experience? If not, it might just make sense to forego storing that data.
- How should I store the data?
Depending on data volume, you might get away with storing in a lighter option like Excel. With more data you might need to set up tables in a cost effective database like Google BigQuery. I generally do not recommend Excel as files can get lost or manipulated. If you have a group of stakeholders that only function in Excel, you could make files available to them while securely backing up data in a database. A visualization tool like Tableau or Google DataStudio could be a great resource here as you could tap historical data from a given database and overlay it with data from Google Analytics for needed comparisons.Our Data Science & Engineering Solutions are always a custom fit for your situation and often the simplest way to get exactly what you need.
Communicate & Educate
During rollout, count on a few die-hard fans of the outgoing analytics system. To reassure any holdouts, build buy-in, and ramp up the rest of the organization, create a glossary comparing and explaining key concepts and metrics like visits v. sessions. Consider how-to documentation, videos, and live training sessions to show users how to pull their favorite reports from the new platform. The educational effort will be time-consuming but will pay dividends in the long run. Down the road, you won’t be peppered with “how do I…” and “can you…” questions. Training your user base is a time savings as much as a quality consideration.
Switching web analytics tools can be cumbersome. Don’t underestimate the effort. For optimal outcome, take the time early on to think through why you are transitioning and how it will be done. Your commitment during the planning stage will be rewarded with more than just a new tool. Going through the process outlined above will greatly improve your analytics implementation in general.
Of course, don’t hesitate to reach out if you are considering a web analytics migration. As Google certified Service Partners, Empirical Path consultants have the experience and commitment to quality needed for a trouble-free transition.