The fun part of being an analyst is being able to work with data, and give insights and recommendations to the key stakeholders to drive business forward. The majority of the business participating in eCommerce are very likely expecting the analyst to come up with things that are ‘Working’ and things that are ‘NOT Working’.
If you’re stuck in analysis top pages viewed, top products sold, overall traffic, then I hope this article inspires you to think in creative way to get into the insights in a structured way.
Here are data points to consider in approaching what is working and not working.
Basic measurement to consider is traffic, conversion rate, and order size. Then measure those metrics against your target or expectations.
Traffic is too general, so analyst needs to look at traffic from different dimension or segments of data. Common area to look are:
- Traffic Source: One of the typical areas are interests are traffic source dimension. Common traffic sources to observe are direct traffic, paid search, organic search, emails, feeds, affiliate network, banner ads, social media, etc. Which digital channels are driving how much traffic, converting by how much, and how much revenue it is contributing to your business.
- Keywords: Keywords from search engines or even on site search will give you a hint on your prospects interests or intent. Is the traffic growing from branded terms or product related terms? To be more relevant maybe your marketing efforts should target specific audience that should land on your site with some product related terms.
- Content: Your site’s content could be a product catalogue or other non product content supporting to drive sales. It is important to know what is contributing to the sales not only from what is selling, but critical to know what content is attributing to closing the sales.
- Segments or a block of data group: After you work out the data exercise of looking at traffic sources and content/products, what unique segments of users are converting? This block of segment of data could be determined by your business’s experience or even effort to drive sales, etc.
For example, if you know people who became a fan or clicked on Facebook “like” buttons are likley to become a customer, then you may want to look at traffic growth for that segment of visitors who are clicking on “like” buttons and becoming a fan of a product or brand. That’s an example, but it is absolutely critical to identify that key segment of traffic that has positive impact to your business.
Segments are really powerful
- Geo location: It is important to think geo location beyond the standard reports in web analytics. Reason I say this is because if you have an offline presence as a store or through products sold through offline retails, then web analytics could potentially become a powerful tool beyond online efforts.
Say you have very weak product placement in one region offline, but you know people from that region makes up a large percentage of traffic online. Then that’s an insight highlighting the actionable opportunities to do something about that area offline. Same goes in the other way around. As more online tools allow marketers to target traffic by geo location, if online presence is weak in area where offline retail sales are strong, then we could potentially be doing more online to target and reach new audience.
- Measure conversion against target. Assuming you’ve identified the conversion segments you’re focusing on to grow your business, it is absolutely important to measure that segment’s conversion’s performance. If it falls short against expectation, then gotta ask yourself why?
- User intent. When you start digging into what’s working and not working, you could be over whelmed with many data. I like starting from known initiatives where a particular group owns that piece of project / feature / function. With that said, dive into those piece of area with data and see what consumers are telling you about the site that causing lower/higher conversion.
One example, is if you’re investing in on site search (like Google Search Appliance) then it is worth looking into what consumers are looking for and how much is that attributing to the commerce conversions. I’m sure there are other efforts that you could start look into that’ll tie the consumer intent to conversions. This type of analysis could be done by looking into segments and diving deeper.
- Cost against efforts. Conversion performance could turn out as increasing or decreasing or performing well or not well. One important thing to think about is how your conversions are turning out against investment (or costs). If you’re spending some money to drive sales, then it is no brainer to look at the conversion performance against your invested dollars. How is that performance per dollar spent looks like? Are you paying too much?
Lastly, looking at the oder size will give you a sense rather you should focus and prioritize your effort against that segmented data or behavior. Say one segment of analysis shows good traffic and conversion, but it only yields an average sales of $5 vs. $100 in other segments. Obviously, that is a point to understand why there is a difference, but it is probably worth prioritizing and looking into what’s broken when the dollar value is high.
Once you start creating a basic practice and matrix, the analysis becomes very fun and requires creative thinking on how you can find that golden nuggets.