Digital Analytics 6 min read
Drive Digital Differently With UsLet's Talk
Here’s How to Grow Your Ecommerce Business With Digital Marketing
- 01 Jul 2022 9 min read
Stockpiles of data can leave you paralyzed rather than acting on business growth. These tips should help keep you moving.
There are about 200 reports in my inbox right now. A couple more stories I can access in a dashboard, but I won’t. All these reports are essential. Lead-to-sale ratio. Average order value. Sales cohort.
Data, your superpower
These are data — useful blocks that, stacked on top of one another, — lead to…wait, what is their purpose again? For increasing sales, I suppose. More data, more sales, isn’t it? All the time, or not at all – if used wrong.
What am I leading to? It’s that making profitable sense out of data and analytics isn’t easy. But at least it doesn’t entail wearing a super-suit or opening the multiverse.
Although it is imperative for you to act on the data you have. Data that can give you an instrumental, competitive advantage. But having a ton of data is only one side of the story. Arguably, it’s also the easier one.
Yet, data, your kryptonite
The – needless to say – tougher one, is understanding this data as you look at it. And, even more so, not getting stuck trying to analyze every bit of it.
1.8 billion consumers are spoiled with choices. That’s data according to a study on online shopping. On top of that, the ecommerce market has around 12 to 24 million stores worldwide. This includes big-name sites like Magento and Shopify, competing alongside third-party commerce. The tough competition that’s getting tougher given the rise of social commerce.
They are the tools of the competition as much as they are your own. They are a source of data you can use in your strategies, too.
Yes, data. Again. Dangerously tipping towards becoming noise. It becomes this when it paralyzes you and hinders action rather than empowers it. When it leaves you stuck because you could be looking at the wrong data, or misreading it.
Innovation – using data to empower
More than ever, it’s becoming crucial for businesses to diverge and innovate. To make sophisticated use of your data to grab the consumer’s attention, and increase sales.
Now, I want you to think of the most recent item you bought online. Clothes? Items from the health and beauty category? Electronics? Home and decor?
If I remember it right, the last time I bought online was a car phone holder. From then on, every time I visit the ecommerce app where I bought it from, all I see are items related to cars. Almost everything!
I see dashcams, all kinds of cords as if each of them varies in function. But a few days ago, I bought 2 cords. I can’t help but laugh while confessing this.
But that’s just me. What about the billions of other consumers seeking to buy and personalized experiences? Imagine these people crowding your ecommerce apps, coming from different marketing cylinders.
As a business, or ecommerce manager, you always face these diverse audiences. They have different preferences and find you from various channels. Imagine the dizzying flurry of data.
I’ll bet, it’s an all too familiar scenario.
Unleashing the power of your data
Last September 26, Propelrr took part in Seamless 2018 – one of the largest marketing conferences in Asia. There, we talked about unleashing the potential of your data.
Before anything else, let me pose a question: How do you handle a crazy amount of data and use it to your leverage? Where on earth would you start?
Putting it in a framework, of course. Into a deliberate and malleable thought process to map out the where down to the how of your strategy.
And, if even with that, you struggle to start, then don’t fret. After all, we’ve perfected a mammoth of a digital framework you can take and use for your own efforts. An effective digital marketing strategy builds itself on top of and around a framework.
In this article, I want to take a moment to zoom into the analytics component.
Businesses these days, regardless of industry, are dealing with more data than ever. And the future we’re collectively heading towards will involve more of it and more experimentation.
Here, moving past thinking in silos is critical to improving consumer experiences. Hence, the need to break the idea of compartmentalizing data as web, CRM, pay-per-click (PPC), mobile, or other.
Ecommerce isn’t so much about infrastructure anymore. Many are able to invest in web design and development already. Heck, we’ve even moved past glorifying data.
The next key thing, if you ask us, is how you alchemize that data – experimenting with and using it to optimize at scale. All to provide the persistent need of customers for convenient and personalized experiences.
This process of experimentation starts and ends with analytics.
Shoprr, the hypothetical client
Take this hypothetical client, Shoprr, as an example.
Shoprr’s bottom line is business growth. They cite ROI as one of the indicators they’re achieving this. They also pinpoint Customer Acquisition Cost (CAC) and Return on Ads Spend (RoAS) as factors that impact this.
These two metrics will then determine the budget to be used, and the channel optimizations to which it is allocated.
Later on, as the data came in, they realized that low CAC and high RoAS didn’t point to growth. They were also not good long-term indicators.
They may have acquired customers at lesser costs and gotten high returns for every penny spent on ads, but overall profitability wasn’t stable. It also relied too much on discount promotions.
The customer journey
To further illustrate, let’s picture the customer journey of this hypothetical client:
Say, Maria, about 28 years old, browses through the Shoprr App and buys two to three maternity dresses. She asks:
- “Will this dress fit me well?”
- “Is this stretchable?”
- “What color would best suit me?”
She’s considering a lot of things.
To help Maria shop, you as the brand should track and monitor her cues as data. Ask yourself:
- “What channels did she come from?”
- “What specific campaign led her to the app/store?”
- “How many items did she purchase? The order value, pages visited, and more.”
After the sale, you can also track different metrics and data like delivery, customer rating, return visit, and repurchases.
If you are this hypothetical client, Shoprr, how would you process all this data to ensure you work your way to grow your business? Are CAC and RoAS enough to predict your business growth?
Moreover, how would you use these data to make sure Maria would be a valuable customer for your business in the long run? Given all the data available, how would you make shopping feel more human and personal for your consumers?
The potential of data
Quite soon, Shoprr started considering the performance and the secondary metrics, predictive indicators, and other non-numeric factors. Through regression, Shoprr was able to identify that it isn’t just CAC and RoAS that determine their estimated ROI but also the Customer Lifetime Value (CLV).
Customer Lifetime Value is the potential profit attributed to the entire future relationship with a customer. As no analytics automatically churns it out, most businesses would need to build their own CLV model commonly using Regression Analysis.
CLV matters because it helps businesses make intelligent decisions concerning marketing, product development, customer service, and potential sales. It helps answer questions such as:
- Marketing: What does my customer acquisition cost look like?
- Product: Which products would best fit my customers?
- Customer Service: What is the cost to engage and retain customers?
- Sales: Who are the customers that need more focus for sales targeting and upselling?
Going back to the customer journey, Shoprr made use of all the captured data to identify Maria’s CLV potential. This led to identifying and strategizing for better opportunities in maximizing ROI potential from her.
Data from previous transactions was also used to plan what ads would resonate more with her. It also helped determine when to retarget the ads, to which channels they are deployed, and what offers would motivate her to purchase.
The end result: Growth
Imagine scaling this to all your online customers. You’ll encourage dozens more to shop while improving their experience of it. This leads to lesser frustrations and more recommendations from your shoppers.
And what will that all result to? Business growth that drives and sustains itself.
Moreover, identifying your CLV would help you save on marketing costs as you focus on high-value customers and improve on other areas such as logistics, which has a greater impact on customer experience.
After the dizzying math, how exactly would you start unleashing the potential of your data?
Steps to unleashing the potential of your data
Many marketers shudder at the mention of ‘math’. But the numerics of data gives you stronger and long-term ROI and so, shouldn’t scare you away. In fact, unleashing your data’s potential takes only six counts in arithmetic. These are:
1. Map out your objectives
Elementary so far, isn’t it? Start with stating what you want and what you’re trying to achieve.
Some guide questions for this:
- What is the primary goal for your business?
- What are the things that could impact your goal?
- How do you measure the said impact?
If you set out into digital marketing and experimentation without this, it could be almost as detrimental as not doing anything in the first place. Moreover, you’ll end up frustrated – wasting time and money over your efforts, or lack thereof.
2. Identify your data models
Now, you lay down some parameters to answer your questions. Without them, you are misled into believing you’re achieving your goals when, in fact, you aren’t.
Guide questions to identify data models:
- What is the main factor that affects my primary goal metric?
- Can I predict my KPI performance? How?
- What is the sales journey of my customers?
How to identify your data models:
- Define the types of information that are needed to be held in the database.
- Identify the relationships between the different data components.
- Pin down anticipated use cases related to the different datasets.
By now, you should have a concept of what you need to create and the types of interactions that are necessary with the data and the database
3. Choose your analytics platforms
These are your tools of the trade.
Guide questions to identifying tool requirements:
- Does it have the features you need?
- Is it easy to learn, use and navigate for all stakeholders?
- What are the costs and the promised value?
Like any other technology investment, these platforms must help speed up the process, crunch tons of data, and support your current and future business goals.
Some other considerations in choosing an analytics platform are:
- Inquire with your internal stakeholders and outline what role they will play in the selection.
- Prepare your data governance plan.
- Identify high priority and low priority features and functions.
- List down at least 3-5 platforms to choose from.
- Come up with a qualification checklist. You can use the criteria below:
- Analytic approach
- Features and tracking types
- Connectivity and integration
- Professional services and support
- Data storage options
- Legal compliance
- Supplier software reliability
- Ongoing and future costs
- Data ownership
- User customization
There are a lot of choices for the analytics platform. I know. It’s both a blessing and a curse. Narrow it down using the checklist above.
4. Identify your analytics models
To put it on record, an analytical model is a mathematical equation that describes relationships among different variables in a data set.
Analytical models are key to understanding your data, generating predictions from logical patterns, and making data-driven business decisions. Without these, it’s nearly impossible to gain insights from data.
To ensure you’re achieving these, here are some questions to ask yourself:
- What report segments do you need to answer your questions?
- How should you present your reports? Dashboard, tables, graphs?
- Will you need to use cross-tabulations or single dimension reports?
How to map out your analytic models:
- Review the objectives.
- Identify hypotheses and prioritize the data sets based on the objectives identified.
- Sift through various data sources to find all the data sets needed. Document each potential data set using the checklist below:
- Access method
- Data characteristics
- Business rules
- Data pollution
- Data completeness
- Data consistency
- Standardize and enrich the data to meet requirements for data processing techniques. Eliminate variables and values that aren’t relevant.
- Deploy the model.
- Track versions, audit usage, and manage a model through its lifecycle
Analytical models are powerful. Imagine being able to make predictions – anticipating what you need to streamline business processes, reduce costs, increase revenue, and improve customer lifecycle.
5. Insights, analysis, and validation
Here we are at the exciting part: Finding out what the data points to and which of it you can put into action.
Here, you mainly ask yourself:
- What stories do you get from your data?
- Were your data able to answer your questions?
- Are you able to measure your performance?
A caveat to note here is that you’ll have to remove your biases. Eliminate them completely, as there simply isn’t any room for false assumptions, especially when your data is proving your ideas wrong.
Maintaining this growth, with a simple acronym:
Quite the contrary, data can outgrow the business, the analytics platform, and the models you built if you don’t continuously strive to improve.
To achieve real improvement, the impact of change must be measured, and the change must be continuous. This makes it possible to determine if you’re going to either survive or dominate or, maybe, fail.
We’re at the precipice of an era where data could determine and destroy the future. More and more, organizations are finding that their economic value is determined by the strength and intelligence with which they experiment and analyze data.
But as some companies are just opening their eyes to this, others are already experimenting at scale. If you find yourself in the former, here are some takeaways that might nudge you to take the forward step:
- Focus is important. The data influx can lead to more confusion than direction. Keeping a careful focus on your goals and business values will help you maintain clarity and alignment. As long as something doesn’t fit either of these, you don’t have to pay attention to it.
- Data is useless without action. This makes it important to have a bias for action. The minute you feel like you’re overthinking something, you probably are. When you’re stuck in this, consciously pull yourself out of it, and refocus on what you can act on, now.
- Your data is only as good as who interprets it. Before putting data and experimentation on a pedestal, remember that the people who are interpreting and conducting them are more important. Invest in upskilling people, and make sure to sow a culture for unbiased data thinking in your organization.
If you think it’s too late to get started on experimentation, well, it’s really not. The driving ethos of experimentation is always questioning the norm and seeing how else you can build on top of them. Those who truly understand experimentation know that they don’t have everything figured out, and so are just as tearing things down and starting from scratch, as much as you are.
Get additional tips on experimentation and unlocking the potential of your data weekly by subscribing to our newsletter.
Choose topics you want to get regular updates on: