Changing digital analytics products within an organization is difficult. Even when organizations don’t feel like they are getting value from their current digital analytics solution, they tend to keep it because switching to a new product often includes the following:
- Reimplementation – Instrumenting a new digital analytics solution normally requires retagging your website or mobile app.
- Loss of Historical Data – In some cases, you may not be able to import historical data which can make it impossible to see year over year trends.
- Retraining – Switching to a new digital analytics product forces you to retrain your internal users on a new user interface.
- Vendor Fatigue – Many organizations have tried different analytics products over the years and in some cases they end up with the same exact issues. In these cases people within the organization believe that switching to another analytics product won’t solve their problems.
These factors often cause a significant level of inertia within organizations that can lead to using a digital analytics product that no longer meets their needs. I believe that there are some cases in which organizations should stay with their current analytics product and fix the organizational issues that are most likely inhibiting success. I have also seen cases where the current digital analytics product is no longer a good fit for the organization. The topic of digital analytics product reimplementation has been topical lately since many major players in the digital analytics industry are encouraging their clients to upgrade to new versions of their own product, which has made organizations wonder if it is a good time to see what else is out there since they have some work ahead of them regardless. In this post, I will tackle each of the barriers to switching analytics vendors head-on and share what I have seen during my career in digital analytics.
Having spent much of my career in the implementation area of digital analytics, I can attest to the fact that reimplementing can be painful. Each digital analytics vendor has its own implementation quirks and even if you implement the same data points there will likely be some tagging updates that are needed to move from one vendor to another.
However, now that most of the major digital analytics vendors have moved to an event-based data model, transitioning from one vendor to another has become significantly easier. In many cases, the actual events and properties that digital analytics vendors require are being collected by a CDP or an internal data collector. Both of these data collection tools can be easily pointed to a new vendor or sent to your current vendor and a new vendor product that you are interested in trying out. Instead of taking weeks or months, it may be possible to send data to a new digital analytics vendor in days or hours.
At Amplitude, we have seen many organizations use CDP products like mParticle and Segment to send us data at the same time data is being sent to the incumbent vendor. This allows our prospects to compare our digital analytics product side-by-side with their current vendor with the same data.
Another way to try out new analytics products is to leverage existing tag management system implementations. For many organizations, they have spent years building robust data layers that have been modeled in their tag management system. This is a semi-vendor agnostic approach to implementing digital analytics. The most popular tag management systems on the market today are Google Tag Manager, Adobe Launch and Tealium. If your organization is using one of these tag management systems, most of your data collection happens in these products and you simply employ tag management rules to send data to your chosen digital analytics vendor. So, for example, if your organization wanted to try using Amplitude, it could use one of our integrations with Tealium, Google Tag Manager or Adobe Launch to re-use existing data layer and tagging work and get their existing data into Amplitude within hours.
I highly recommend periodically trying out new digital analytics products alongside your existing one before completely switching to a new vendor. Doing this will allow you to be sure it is a good fit for your organization before you cut off the incumbent vendor and will start to build up some historical data in the new product.
Loss of Historical Data
When considering a digital analytics vendor change, I often hear the loss of historical data as a major barrier. As an analyst, I can definitely appreciate the need for historical data, especially for organizations that have data seasonality. However, in my experience, the most impactful analyses are those that are focused on what has happened in the last few months and what might happen in the future. In practice, I have only leveraged year-over-year data for about 10% of my analyses.
But if historical data is important for you, here are some suggestions I would have when considering switching digital analytics vendors:
- Overlapping Products – If your organization can afford it, you may want to use the old and new analytics product concurrently for a year. This will allow you to build up enough historical data in the new vendor product, while also providing the historical data in the old product. Obviously, not all organizations can afford to pay for two vendors, but if historical data is important enough to justify it, this is the easiest solution.
- Leverage Data Warehouse – Many organizations are sending digital analytics data into their own internal data warehouse. This is done to keep all historical data, especially now that many vendors have begun removing data after 25 months by default. Digital analytics data is also sent to internal data warehouses so that it can be combined with other customer data, such as call center data, point of sale data, etc. Since historical data exists in the internal data warehouse, it’s typically possible to leverage that for the rare occasions that you need to compare different time periods until your new digital analytics vendor has sufficient historical data.
- Backfill Historical Data – In many cases, it is possible to backfill historical data from your current analytics product into your new analytics product.
If one of these are an option for your organization, the dependence upon historical data can often be removed as a blocker for trying out a new digital analytics vendor if your current vendor isn’t delivering value.
When you consider switching digital analytics vendors, retraining current users should definitely be one of the considerations. If your organization has 500 users that actively login to your digital analytics product, training them on a new product can be a challenge. However, I have found that in reality, there are many fewer people than you think who are actively using digital analytics products. Even at large organizations, I have seen many examples where a core digital analytics team makes up 90% of all digital analytics product usage. So the first thing I would do is identify how many people are really consistently using the current digital analytics product today. Most digital analytics products have a way to view this natively within the product, or you can purchase add-on products that will provide this information.
But if you really need a new digital analytics product, you just have to bite the bullet and retrain people. Sometimes the new digital analytics product may have an interface that makes it easier for your internal users to learn. Even though it is painful, I have found that sometimes retraining internal stakeholders is a good opportunity to reconnect with them and find out what is important to them now and what questions are top of mind. A large portion of success in digital analytics is relationship building, so instead of looking at retraining as a chore, consider looking at it as a great opportunity to build deeper relationships with your internal stakeholders.
Many organizations have been doing digital analytics for twenty years or more. During this time, most have cycled through multiple vendors with varying degrees of success. I am a big believer that any organization can be successful with any digital analytics product, since many of the things that cause organizations to struggle with digital analytics are product-agnostic. If the analytics product you select is the determining factor as to whether your organization is successful or not, you likely have other problems at play.
When I was an analytics consultant, I once had a new client kickoff call where the client started challenging me on my approach to the new implementation. This client, without trying to be ironic, said to me, “We are going to implement it our way…this is how we did it with Webtrends and Coremetrics and how we will now do it with SiteCatalyst…” I was so annoyed that I inadvertently blurted out, “…and how has that worked out for you so far?” With this client, I could tell immediately that the analytics product was not what was causing them to fail and that without making structural changes, they would fail again with the new product.
All this being said, there are times when, despite possible vendor fatigue, organizations should consider changing digital analytics products. Here are some of what I would consider to be valid reasons:
- Customer Behavior Changes – There are cases in which your customers make significant changes in their behavior that can impact your digital analytics vendor choice. In the past decade, the most significant change has been the migration away from desktop websites to mobile devices. For many organizations, mobile app activity was 5% of their digital interactions a decade ago, but is 85% today. In this case, if there are digital analytics vendors that provide better insights for mobile apps than the incumbent analytics vendor, it may make sense to revisit which vendor should be used going forward.
- New Team Ownership – Over the years, digital analytics has been owned by many different teams. For many organizations, the initial ownership was the marketing team since digital analytics began as a way to judge the effectiveness of digital advertising spend. But as digital experiences evolved, in many organizations, the ownership of digital analytics has shifted to either a centralized analytics team or a product management team. If digital analytics moves from one team to another within your organization, in my opinion, the team inheriting the analytics function is justified in reassessing whether the product being used is the one they want to use in the future.
- Fresh Start – As mentioned above, I often find that problematic organizations use switching analytics products as a way to gloss over real organizational issues. But sometimes, the digital analytics function at an organization has gotten so bad that it needs a fresh start. For example, when I joined the marketing team at Salesforce, the existing digital analytics program was not in a good state. In my first few weeks, I interviewed many stakeholders and learned that none of them trusted the data in the analytics product and that the model being used to support the organization wasn’t working. I knew that the analytics team had passed a point of no return and that a fresh start was needed. In this case, since my assessment showed that the issues were more implementation-based than product-based, I decided to temporarily shut down the analytics product and reimplement it from scratch using my proven implementation methodology. But I could easily envision a scenario where implementing a new analytics product would be a symbolic way to show internal stakeholders that things were changing and would be better going forward. Ideally, the new analytics product would have to be coupled with new processes and anything else that was causing issues with the old product.
These are just the most common reasons that I have seen organizations justify switching vendors, even in the face of very real vendor fatigue.
As stated at the outset, changing digital analytics products is difficult. Choosing which digital analytics product your organization uses is a major decision with many tangential implications. Hopefully some of the items above provide you with some insight into the factors you should consider when making decisions around possibly switching digital analytics vendors.