How to use data-driven insights to optimize change in your organization: 10 expert tips
1. Build a solid data foundation first
Before investing in cool things like AI, machine learning, or a sophisticated digital experimentation framework, get the basics right, according to FairPrice’s Vasu Venkataraman.
The basics of data look like this:
- An updated data dictionary – documentation keeping track of what’s in a database
- A central customer data platform to keep teams on the same page
- Clean data that ensures trustworthy insights.
While this is ‘unsexy’ work, building a solid foundation is crucial for setting your organization up for success in becoming a data-led organization.
Application step: Use the framework above to understand your data and product maturity gaps.
2. Get teams on the same page with a North Star metric
Aligning your team efforts to metrics is vital, especially in complex businesses with many teams, channels, and verticals. How can you ensure that your end goal is the same?
The solution is to define a “North Star” metric that defines the critical measure of success for a product team and unifies a team around a common goal.
A hypothetical example of a North Star metric for a grocery app.
By identifying a North Star, teams can prioritize their focus while avoiding wasting time on vanity metrics or projects that don’t directly improve customer satisfaction or positively impact the business.
3. Look for hidden opportunities to delight your customers
FairPrice’s digital team learned this first-hand.
According to their heatmaps and product analytics tools, people were going to their Accounts page to check on their orders a day before delivery. They also repeated this process after a delivery to ensure they received all the items they ordered.
These hidden insights helped the team simplify a tedious experience by enabling users to access their live orders straight from the home screen.
Takeaway: When planning your customer journey map, there will be non-linear journeys and ambiguous insights. As you review your analytics, look for hidden opportunities to delight your customers.
4. Turn your insights into action with an experimentation framework
How do you build a case to take action on your insights?
Vasu’s answer: start small with digital experiments.
He used his 5-step experimental framework to explain how he would improve basket sizes on a grocery app.
Step 1: Identify your North Star metric and goals. Example: increasing the average basket size per order.
Step 2: Define your hypotheses. Using the grocery app example, ask yourself when customers would be open to a product recommendation and compare the merits of different product recommendation styles.
Step 3: Prioritize hypotheses based on confidence level and effort estimates.
Step 4: Test your hypothesis with a week-long experiment.
Step 5: Analyze findings and repeat
Using this method, you can justify further investments of resources and time based on learnings from multiple smaller experiments.
5. First-party data vs third-party data is more critical than ever
According to Salesforce, 36% of businesses will invest in a first-party data strategy over the next two years, which involves collecting data directly from your website visitors with their consent.
Collecting first-party data efficiently requires trust. Trust comes from showing that your brand protects consumer privacy while providing value and building meaningful customer relationships.
Your next step: Consider a direct-to-consumer (D2C) approach if you work with multiple consumer brands or retail partners and get to know your end-consumers. Check out the other tips to build a data infrastructure that helps you make informed decisions.
6. Keep a historical record of your past performance
In Figma, an interface design tool, R.M.Williams tracks past homepage designs and the week-to-week performance of each page element to help both digital and creative teams understand which creatives worked best.
Tracking past performances meticulously helped their digital team:
- Understand homepage performance on different devices,
- Zero in on what customers found frustrating, and
- Prioritize optimizations with the biggest impact.
Compounded over time, these insights led to R.M.Williams becoming a gifting destination for customers. The team was now ready to create a series of gift guides that met their customer’s needs, resulting in a winning campaign recipe for the business.
7. Not every test needs to be huge
At Australia’s financial services and insurance provider Suncorp, they tested if their visitors wanted to see customer ratings on their web pages, a simple test that could be done in minutes.
Removing customer ratings from their web pages decreased quote start rates by 2.3%, showing displaying social proof helped build trust with web visitors and nudged people to reach out for an insurance quote.
What this means: You don’t always need to think big to drive change with data. Small tweaks can generate the desired results or form a foundation for future projects.
8. Come up with optimization ideas by analyzing data
Australian TV company Foxtel used behavioral data from Contentsquare to develop optimization ideas, systematically designing a hypothesis and recommendations.
Here’s how Foxtel’s Hunter He builds a recommendation with data.
Observation: Foxtel observed higher-than-average click rates on their Popular Questions content in the footer of their order summary page.
Hypothesis: Many customers had to scroll to the footer to find important information. Therefore, making it easier for customers to get this information could improve checkout numbers.
Recommendation: Get people to the content they want to see. One possibility to test is to add an anchor link to guide the users to the FAQ section at the bottom of the page.
9. Reflect after each iteration
Not every digital experiment or iteration will be successful, and that’s fine!
Besides analyzing data from your analytics tools to determine if an experience is successful, have a system to reflect on each iteration.
Foxtel’s Hunter He turns to these 3 questions to reflect on any unsuccessful iteration.
- Is the problem you’re solving an actual problem?
- Was the idea or hypothesis incorrect?
- Was the experiment executed poorly?
10. Share your successes and failures with a professional community
Driving change with data is a journey with ups and downs.
Decathlon’s Roanna Zheng shares that “having a like-minded professional community to share successes and learnings was essential for us.”
We’ve got a bunch of ways for you to find your (CX) tribe. This could be attending one of our flagship CX Circle events, learning from other CX professionals in the Contentsquare community, or empowering your team with Contentsquare’s training and certification program.
Missed it? Don’t sweat it.
If you couldn’t attend our CX Circle events, don’t worry because we’ve got you covered. Watch our CX Circle events on-demand here.
Keep Reading...
