In the first part of this series we discussed the following factors that indicate the need for a transformation of your digital analytics.
- Translation of business needs into analytical use cases
- Getting actionable insights from your partners
- Comprehensiveness of digital analytics
- Cost of digital analytics platforms
If you haven’t gone through the first part yet, it might be a good idea to read it here
In this part we will discuss the following indicators of your digital analytics transformation.
- You are forever in the process of implementation
- Your analytical tools and reports are siloed
- You are often worried about reporting accuracy
- You struggle with optimization
- You are worried about the transformation project timeline
#5. You are forever in the process of “implementation”
Are you a marketer who has already started an ambitious digital analytics journey, but is unfortunately stuck midway and is fighting to get your analytical implementation right? Then, this is just for you.
You were floored by the range of features that your analytical tool offered.
You might have had a clear idea of what you hoped to achieve.
It was your business need / goal that had driven your analytical roadmap.
But, seeing the range of features in your analytical platform, you wanted to leverage everything you thought was “cool”.
As a result, you not just lost track of your path to achieving your original intended business goal, you haven’t really cracked the cool features either.
By the time you realized this, you crossed the original project implementation timeline by a significant number of months.
As damage control, you start all over again, strategizing to achieve the original goal and the implementation revision would go on for months giving you a feeling that you are forever implementing the analytical platform.
This is true of many marketers wanting to leverage all the features in a single enterprise analytical platform.
Imagine how long the implementation would go, if there were multiple platforms at play.
You are not alone. There are thousands of marketers like you.
In fact, we have quite a few clients who faced this challenge before they came to us.
One of our clients, a Malaysian Bank had the same problem. They were working , not just to get their implementation right and freeze on right metrics, but also on saving costs owing to a large number of inappropriate server calls.
We intervened and optimized the implementation in under 90 days, standardized the metrics to be captured and also integrated the analytics platform with targeting platform to provide highly personalized experiences to their audiences.
Regardless of whether you are starting your analytical implementation afresh or are already in the process of implementing forever, here’s what you should be doing.
- Clearly define your expectations from the analytical tool
- Most importantly, “Define Project / Implementation Exclusions”
- Prioritize the list of features that you want to leverage post implementation
- Ensure that the solution definition is flexible enough to accommodate future needs without compromising on the implementation timeline
#6. Your digital analytics stack is siloed
Your analytical maturity might be reasonably high.
You may be tracking your marketing website not just for traffic, but also for behavior and conversion.
You may be tracking your mobile app, transaction portal and digital campaigns as well.
More importantly, you understand the need to provide personalized experiences to your audience and have a personalization platform in place.
Source: Salesforce State of Marketing Research 6th Edition
Despite all this, your audience engagement and conversions are dwindling.
What could have gone wrong?
Well, this is apparent. You are collecting and acting on data in a siloed manner.
It’s not just you. A lot of marketers struggle with this problem.
You may be providing new channels to your audience to engage better and may even have a great adoption rate.
When your audience interacts with you across multiple channels, they just don’t move between channels, they also move back and forth in the purchase journey.
This is exactly where the problem lies.
When you don’t have integrated analytical systems in place, you track audience interaction and respond without considering how the audience engaged on other channels.
As a result, your communication loses context and results in lower engagement.
This is what differentiates analytically mature and successful brands from others.
Analytically-mature brands don’t stop at collecting multichannel data, they also integrate those and intelligently use it to maximize context relevance and deliver hyper personalized communication that results in greater engagement and conversion.
Unfortunately, the number of brands that do this are very few.
One of our clients, India’s most diversified NBFC that offers instant approval of loans is relatively mature vis-à-vis other players when it comes to digital maturity with stellar websites and mobile apps. Their digital budgets are pretty steep with sophisticated platform subscriptions involving analytics, targeting and data management platforms.
However, the client was using all these platforms in a siloed manner and was not able to target visitors with personalized content.
We started implementing the rollout of the targeting platform on iOS and Android mobile apps (which was missing earlier) with proper server calls without affecting the page performance, which was key to their business.
To ensure visitors were profiled correctly to maximize targeting relevance, we integrated analytics, advertising and targeting platforms and combined first, second and third party data and arrived at 100+ complex targeting use cases not only based on the online behavior but also based on their attributes like Age, Gender, and Occupation etc.
These were tested and deployed on the portal and mobile apps on a regular basis and ensured relevance of communication.
Being an established brand is very different from being analytically mature. If you want to transform your current analytical capability to be at par with analytically mature brands that boast of their capabilities to be able to provide hyper-personalized experiences to their audiences, day in and day out, in their investor reports, interviews etc., here's what you need to do..
- Define and map customer journeys for the products and services that you offer. Ensure the journey extends beyond purchase / sign-up to ensure audience intents are captured comprehensively
- Determine what events (clicks, taps, scrolls, focus fields etc.) in each of your touchpoints such as marketing & transaction websites, mobile apps and other digital channels represent what intent
- Integrate platforms tracking these events across touchpoints to establish the right context and determine the next best action for each user
- Present the next best action to the user regardless of what channel he / she visits next
- Track the user’s response to your communication and optimize the same by trying variants of your communication
#7. You are often uncertain / unaware of reporting accuracy
It is not easy to find a marketer that says the report that he / she has received is perfect and in sync.
You would have come across such discrepancies when you have reports coming from multiple platforms beyond your web analytics tool such as your CRM, CMS, Personalization / Testing tools etc.
Here are some of the common reasons why such discrepancies arise.
- Data - such as number of leads generated – is often tracked by analytics tools based on the number of times the ‘thank you’ or the confirmation page is loaded. If your users are able to load these pages manually and if they happen to load it multiple times, your analytical tool is likely to show a high number. Remember to provide a dynamic or unique thank you / confirmation page for each successful form fill / transaction etc.
- Your team tests stuff during production. Again, every single marketer would have encountered this scenario. Most times, you want to ensure your audience will see the right pages and your end up testing on the live website. Regardless of why, it is a bad method as it impacts reporting data. To overcome this, either you need to restrict your tests to Test URLs or you need to exclude IPs that are used for testing.
- You are comparing the metrics at the wrong time. If your analytical systems do batch updates of analytical data, you may find some discrepancies if you are comparing before the scheduled update. Similarly, if your analytics tool reports only sampled data, discrepancies are bound to happen.
- If you compare data from two different analytical systems with different definitions for same metric, then discrepancies are likely to occur. Always ensure the comparison is done only on metrics with the same definition.
These are minor discrepancies that can be easily rectified.
However, if you aren’t hundred percent sure that the data that you are getting is accurate, that’s where the problems arise.
The actions that you take based on such analytics metrics may lack relevance and will result in lesser engagement and conversion.
Usually, analytical tool implementation is done by one partner and analytical reporting and ongoing one-off tagging are handled by different agencies.
If the solution definition best practices are either not handed over to the reporting agency or the reporting agency stops following the solution design considerations, over time, data discrepancies are very likely to occur.
Our client, the world’s largest semi-conductor manufacturer had just migrated from an enterprise class analytical tool to a free analytical tool. The onus was on us to replicate the rich analytical reports which were available earlier with the enterprise analytical tool.
Since we were already delivering analytical reports, we had a good understanding of their business needs and what data was important to aid in decision making. But we had a great challenge ahead of us as the free analytics tool that they had chosen was used to give only sampled data. Any significant difference in the data that get reported would be seen as an indicator of our competence.
We needed to overcome it and provide a greater emphasis on solution design to ensure the desired data was tracked. Since the client’s portal educates developers, it had a lot of content and tons of new content gets added every day. To ensure tracking is done correctly in new pages, we had developed a tagging checklist that ensured desired metrics are captured consistently. Once we ensured this, we were able to deliver insights of same quality as earlier.
Sadly, most of the marketers are unaware of this discrepancy and its detrimental impact.
If you want to be sure that the data that is reported is as close to eality as possible and your actions based on analytical reports, as relevant as possible, here’s what you need to do.
- Ensure the recommendations provided as part of solution designing percolates to the day-to-day activity level
- Create a checklist of sorts covering nomenclature, query strings, optimal server calls, events etc. to be followed for each tag
- Mandate your QA team to validate that the solution design recommendations are followed whenever new tags are created or existing tags modified.
While these may appear too operational and not a marketer’s cup of tea, this impacts marketing costs and outcomes directly and you need to ensure this.
#8. You don’t focus on optimization enough
A typical analytical journey comprises the following activities.
- Discover: Identify business needs that can be fulfilled by analytics
- Define: Design solutions to come up with tracking codes to collect metrics that impact the intended business goals
- Collect: Deploy the tags across all touchpoints and start collecting data
- Analyze: Analyze the collected data and deliver actionable insights
Most marketers get a variety of digital analytics reports and they make one-off decisions / interventions based on those.
However, apart from the above 4 activities, there is one ongoing tricky activity that most marketers struggle with. That’s optimization.
The term optimization is too generic to mean anything. What do you need to do as a marketer to optimize your analytical efforts?
Ideally, optimization needs to be done at two levels.
- Optimizing tracking codes
- Optimizing customer experience
Optimizing tracking codes:
The ability of your analytical platforms to track user behavior would have grown significantly in comparison with what it used to track when you initially implemented it.
If you have to leverage those capabilities, you have to keep improving the way you track your audience behavior and the only way you could do is by optimizing your tracking codes.
These are some of the parameters you can optimize in tracking codes.
- Number of server calls
- Range of variables used
- Range of events captured etc.
Optimizing customer experience:
There two ways marketers initiate optimizing customer experience. Some marketers use heuristics and their creativity and some use analytical insights.
While it’s okay to use heuristics to optimize experience, more often than not you will not be certain about the outcomes it may produce.
That’s the reason why some of your interventions succeed and some don’t.
Whereas if your interventions are triggered by analytical insights and you couple that with your creativity, that’s when you achieve resonance with your audience needs and eventual improvement in conversion.
There is a large foreign bank in India for whom we are managing their digital presence as well as help with analytics. During the early years of engagement, we used to handle both these functions in isolation and were struggling to meet leads targets. We used to make random changes to the portal by using our creativity to maximize lead generation.
Over time, we realized this isn’t sustainable and that’s when we decided to tag every inventory in the portal and statistically arrive at benchmarks for each inventory based on past performance. We made it a process inside that regardless of leads pressure, we will not make changes to any inventory unless there is a dip in performance over seven days. This turned out to be master stroke as we are regularly able to meet monthly leads targets and also provided us utmost visibility to plan our interventions to achieve leads targets. Today, every single change in the portal is triggered only by analytical insights
Optimization is a largely under-utilized, yet important technique in digital analytics. If you want to maximize predictability of the performance of your interventions and move to the next level of analytical maturity, here’s what you need to do.
- Mandate quarterly review of your current tracking ability to meet your growing business needs and plan for changes, where required
- Mandate that every action in your portal is triggered by an analytical insight and is not based on heuristics
- Test the interventions on a sample of audience before a full blown rollout to maximize success
#9. You are obsessed with tool metrics
We all know digital analytics is about numbers and metrics.
But, what if I say you should not be too obsessed with tool metrics?
Some of you might think this is hypocrisy.
Let me explain.
You would have heard about vanity metrics. Yes, those metrics that don’t really make any sense. For example, hits, clicks, impressions, views etc.
But, in reality there are no constant vanity metrics.
“Any analytical metric viewed in isolation without applying the context is a vanity metric”
And when you are too obsessed with those metrics you end up travelling the wrong path.
Let’s see a couple of examples.
We all know how bad an 80% bounce rate is. It’s 2X more than the average bounce rate.
If we apply a bit of context over it and find the 80% bounce is happening on a landing page, you should be doing something right. You’re not allowing your audience to move elsewhere before converting.
Next, we are all obsessed with the metric, Time on Site.
We assume, higher the metric is, higher the engagement is. An isolated view of this metric is utter nonsense.
If yours is a transactional / functional website with lot of call-to-actions and your audiences is spending a lot of time on the website but the page becomes the top exit source, is it a happy thing? Shouldn’t you be worrying that your audience is not able to find what they want?
There are so many metrics like that you need to be wary of.
Despite all the powerful features of digital analytics platforms, you need to be cognizant that they are not panacea for your business problems.
You need to define what metrics are worthy for you and how you need to look at those metrics.
For example, here are some of the metrics that are worthy in my opinion.
- Entry points by source and volume
- Retention by source
- Profit/Revenue/Conversion by source
Looking at the above metrics, it’s apparent not all those are direct analytical tool metrics. Instead, you need to derive.
By adding an additional parameter such as source, you are able to establish context and meaningfully make decisions.
Looking at right metrics with context is only a start. The real value lies in your ability to take marketing actions based on insights.
However, not many are able to master this.
90% of marketers don’t know what they don’t know, creating problems with consistency and marketers’ ability to act.
End of the day, we are all marketers. Here are some of the marketing actions that you can take after taking a look at the right metric in the right context.
- Re-allocate the media mix
- Micro-segment the audience
- Test different drivers of interest for each segment
- Test effectiveness of niche acquisition platforms
- Run marketing experiments
These are only examples and are not exhaustive. Only when you make the right marketing intervention as a response to the analytical insights that you receive, you would be able to move up the analytical maturity ladder and most importantly, grow your business.
One of our Malaysian telco client with ambitious digital transformation goals engaged with us to help them transform their digital analytics as they were not utilizing their existing analytics platform completely. In fact, rarely they were able to action on the reports that they were receiving.
We started off reviewing their existing reports and realized most of the reports are vanity metrics with no actionable insights. Further, the metrics that they get for their digital campaigns were incomplete and interrupted. As a blessing in disguise there weren’t any action taken based on these reports. Otherwise a lot of media money would have been wasted.
We came up with a list of metrics that were highly relevant for our client’s vision (a lot of those were derived ones involving multiple metrics). We also had to optimize the tracking codes since the existing ones were insufficient to capture intended metrics. Since our client was not used to make marketing decisions based on data, we included exhaustive actionable insights on each report shared. As a result, our client was able to optimize their media mix, discover their audience much better and achieve portal targets.
Here’s what you need to do to ensure your focus is on right metrics and truly leverage the power of digital analytics to drive your marketing.
- Shortlist metrics that directly represent the growth of your business.
- Arrive at how those metrics can be derived by combining metrics that are available in your digital analytics platform.
- Validate data collection method for these metrics and ensure tagging is proper with appropriate events captured and reported
- Rollout data collection, reporting and actioning
Despite using digital analytics for a considerable duration, many brands still have a huge scope for transformation.
A large part of the credit grows to the ever-growing capabilities of analytics platforms, the proliferation of surround systems such as targeting platforms and the deep integration between surround systems of same parent company.
While it’s always tempting to use every new feature that comes with your digital analytics tools, you need to ensure the selection / usage of tools is only driven by your business needs.
Only then you would be able to tame the analytics beast.
Trust you find the points discussed in the series will help you evaluate which part of your analytics need transformation.
Leave your valuable comments below.