Ecommerce Marketing KPIs #ecomchat – Key Points

This week’s #ecomchat was on Ecommerce Marketing KPIs – a subject so intrinsically interesting that people often gaze up to the stars hoping for divine insight.

The chat was based around 3 key question as usual and below is a summary of the key points and discussion threads.

Question 1 – What are the most important KPIs for ecommerce marketing and why?

• @jamesgurd kicked the session off with the durable Per Visit Value, pointing out that you can drill down into the data by looking at it across segments, for example new vs. returning customers and PVV for non-bounce visits.
• @paulnrogers chipped in with conversion rate by channel
• @irakates mentioned ROAS for paid advertising including PPC. @jamesgurd added that including ROI for paid media is really important as with large catalogues product margin can vary wildly, so profitability is an important distinction.
• @danbarker played Devil’s Advocate with PVV and conversion rate, explaining that if you spike traffic really high, then both will inevitably fall. However, lower doesn’t always mean worse, you need to look at the wider picture and it all depends on what your goals are.
• @DanielJTruman suggested Page Value, so you can look at which pages are contributing the most across all goals and then learn from them to replicate success.
• @ChrisLDyson asked if anybody measured customer feedback scores – something that can be really useful when integrating customer survey data into web analytics tools.
• There was a general consensus that KPIs are effective when they are evaluated in context. For example, an increase in a conversion rate by itself doesn’t give you a clear picture, you need to dig deeper to find out what caused the change.
• @chrisdalrymple explained that conversion rate must be split by channel otherwise it’s meaningless. @EdwinBongo added that it is only really useful when you drill down deep enough to have consistent traffic type e.g. keyword level.
• A great comment from @EdwinBongo summarised a lot of the discussion about what makes a good KPI – “if you’re looking for a flawless KPI you’re going to be very disappointed”
• General agreement with @jamesgurd comment that KPIs are really reporting, not analysis, and that they act as an alert system to prompt further analysis.
• @DanielMColeman made the interesting point that if you’re running a large team, you’re probably not going to be drilling down into detail yourself but the KPIs can act as a trigger for delegation.
• @danbarker proposed a basic Topline KPI set: # Visits, # New, Transactions, ATV, Revenue, Customer Satisfaction, Marketing Spend.
• @StokedSEO finds that the number of pages receiving at least one visit direct from Google is a nice one when focusing on SEO.
• @EdwinBongo suggested looking at the average return per new customer on a first order basis, split by marketing channel..

Question 2 – Any good examples of KPIs? e.g. ones applicable to all businesses, surprising odd examples from particular businesses

• @ChrisLDyson proposed changes in shopping cart abandonment as this can be a good indicator of UX issues.
• @danbarker once worked for a place where they had a warehouse metric ‘metres per order’. Used to keep order picking fast.
• @mattycurry likes using fracture rate – the number of products where at least one variant is out of stock. @DanielMColeman uses an event in GA to monitor out of stock views.
• @danbarker mentioned return rate by product as this helps identify product related issues.
• @paulnrogers brought up the importance of using call tracking and @EdwinBongo added you should measure by traffic source/major keywords as well – the learning is don’t undervalue channels because you’re not tracking all the orders. @danbarker indicated you can measure this simply using GA by using a href=”tel:xxx” with an event firing onclick/onmousedown

Question 3 – Attribution: Holy Grail, Red Herring or Other?

This was a quick fire question with only 10mins on the clock remaining….

• @paulnrogers made an interesting point that attribution can be really good for internal PR and general happiness – showing other teams how they contribute to the overall revenue.
• @mattycurry thinks it’s valuable, giving the example that for high-cost generic terms in PPC, sometimes attribution from other channels is the only way to justify spend
• @danbarker said that he finds a lot of people use attribution to help rationalize marketing spend that intuitively works.
• @paulnrogers explained that it can be really useful for agencies, especially when digital campaigns are generating offline enquiries – needs systems/data integration to ensure there is a full picture. He has tried freebies / discount in-store with discount code to attribute online sales to offline and similar the other way.
• @walkerama summed it up nicely – “so with attribution the question is what accuracy is useful for your business and how much work/cost is put into getting data”.
• @therustybear made an important point – it all depends on what business questions you are answering.
• @danbarker closed out with the immortal words, “I saw @EricCrossfield do a talk once, someone asked him how you perfectly join together all channels. He said “Impossible. How do I know if someone walked past John Lewis window on the high street?””

Thanks again to everyone for joining in, we hope to see you next week. And please let us know what you think of the latest trial time of 7pm UK time.

Cheers, James & Dan