The Complete Guide to Retail Analytics
In 2019, the pace and scale of retail has never been greater.
With more and more D2C brands sprouting up every week trying to steal market share from your space, have you ever wondered, what are they all doing to grow so quickly?
Your initial answer might be “customer service” or “venture capital dollars” but that answer is incomplete.
Want to know the answer?
The first order of operations (after validating that you have a product or a brand that shoppers buy into) involves laying the analytics infrastructure to start measuring differentiate initiatives.
- Nobody hires 100 customer success reps without first understanding the cost structure and the expected increase in customer lifetime value.
- No venture capital is willing to continuously invest business that can’t quantity the return on expected dollars spent.
Head to these post directly for shorter reads
Now that you have a rudimentary appreciation for the value of retail analytics, let’s break down the sections (feel free to skip directly to those sections that are more relevant):
- Chapter 1: What is retail analytics?
- Chapter 2: What jobs can analytics help retailers to solve?
- Chapter 3: Data requirements to get started
- Chapter 4: Difference between executive, managerial and operational dashboards
- Chapter 5: 203 metrics for retailers and brands to track
- Chapter 6: 9 metrics for start-ups and small businesses should focus on
- Chapter 7: Operational Analytics highlight
- Chapter 8: Leveraging Analytics for Long-Term Success
What is retail analytics?
Retail analytics is the practice of leveraging data to measure performance, augment decisions and monitor results. In retail, common applications include; measuring the performance of marketing campaigns, monitoring the supply chain, inventory management and merchandising.
A common end result of analytics is the tracking and monitoring of Metrics and Key Performance Indicators (KPIs) which are then employed as proxies to measure the health of the business.
Where does the data come from?
Given the proliferation of technologies today, data can come from many sources. Some include:
- Daily transactions
- Day-to-day Operations
- Shopper Loyalty Initiatives
- Product Information
- Marketing Attribution Tools
- External data sets (e.g. Weather)
New technologies have allowed us to augment our analytical prowess through the combination of data sets (e.g. promotions effectiveness derived from Marketing Attribution and Sales Data) to paint a more complete picture before making a decision.
With the proliferation of sensory (e.g. IoT devices) and the reduction in cost for storage and compute in cloud technologies, it is now easier than ever to collect, store and analyse data.
The goal post, however, hasn't changed.
What should be my next action to help me improve my business by either helping me increase sales or decrease costs?
What can analytics help retailers solve?
TLDR: NOT everything.
You were already doing the jobs that needed to be done before you implemented analytics. Analytics is meant to help you measure, analyze and focus your efforts.
Example 1: Planning merchandising
- “Guesses” as to which product would do well.
- (If you had more than 1 sales channel, it would be too complex to manage the differences)
To provide a more holistic estimate, use;
- historical data,
- product attributes,
- price elasticity,
- traffic forecast,
- customer acquisition cost and
- marketing budgets
Example 2: Streamlining shopper engagement
- Send email marketing to all 1,000,000,000 shoppers on the newsletter subscriber list until something happens
- 1st: Segment list based on open rates (throw away those that never open)
- 2nd: Segment list based on purchasing history
- 3rd: Use product affinity analysis to see which products tend to be bought together
- 4th: Send emails based on the different purchasing history
Example 3: Optimising digital marketing dollars spent
- I know 50% of my marketing budget is wasted. I just don’t know which 50%.
- 1st: Monitor customer acquisition cost (CAC) across advertising channels
- 2nd: Drill down to the Google analytics multi-channel funnel
- 3rd: Reallocate budget from channels that are useless to those that are working
Running a business is f***ing hard.
We don’t always have time to search for the next great analytics tool so, when it comes time to invest in one, it usually falls into one or two buckets;
1. Business is going amazingly well, it is time to optimize
The car seems to be working, time to wisely invest in the engine.
How do I turn $1 into $5?
Common Questions at this point include;
- How do I replicate this success with another product?
- How do I replicate this success with new shoppers?
- How do I optimize my marketing funnel to bring new shoppers?
- What should I do to increase the lifetime value of my loyal shoppers?
2. Business is going terrible, what should I do?
The car seems to be breaking down, why?
How do I turn stem the bleeding?
Common Questions at this point include;
- How did this all happen?
If you fall into the second bucket, I’m sorry to say but it might be too late.
Running a business is hard.
Why not spend a little extra time investing in analytics early so that you won’t accidentally screw it up later?
Here is the reality:
- If you have collected NO data, tracked NOTHING and monitored NOTHING.
You have no way of improving your business, because you have no idea what was going on (besides that there was a fluctuation in sales during seasonal periods)
Assuming you invested early. What can you expect from analytics tools?
1. Augment decision-making in the company
There should be less guesswork in the business everyday that passes.
With more data points around product, shoppers and sales, you SHOULD have a clearer understanding which initiatives should be cut and which are worth investing in.
With the same data set in-front of everyone in the team, EVERYONE should be able to make the same decision (most) of the time without you having to intervene in every decision.
2. Monitor metrics and report KPIs
Bring together a holistic dashboard that can be used to organise and monitor information about everything that is going on.
Some BASIC goals that are worth monitoring include...
- Sell-through rates
- Promotional performance
- Basket compositions
- Pricing effects
- Return on investments in marketing
3. Optimise opportunities
- Funnel analysis to improve the number of completed transactions
- Improve social media ad set targeting to decrease acquisition cost
- Promotional mechanics to improve average order value
- Shopper newsletter to increase lifetime value
- Birthday coupons to increase shopping frequency
If you skipped everything else above, focus on this last section.
How to get started?
1. Write down the functional areas in your company. Some examples
- Customer support
- Digital marketing
- Inventory management.
2. If these areas where 10x better than they are right now. What would that look like?
- Lower average first response time
- Lower customer acquisition cost
- Higher average margins
- Higher sell-through rates
3. Those are your metrics. Measure EVERYTHING that you think might help improve those numbers. Find tools that help you improve those numbers !!!
Read the next section about data requirements.
Data requirements to get started
Read the last section before skipping here.
You might already be sold on the promises of retail analytics. But is your organisation ready?
Do you have the answers to these three questions?
- What problems do I want to solve?
- What sort datasets do I have on hand?
- What is the general level of expertise of my team?
Think about what can be improved in the organisation.
- What if sales was 10x better?
- What if marketing was 10x better?
- What if product was 10x better?
- What if customer support was 10x better?
How do you measure each of those functions?
Let’s take customer support as an example. (commonly overlooked but it could make a HUGE difference)
Problem to Metric
Why does this function exist?
Because it improves customer lifetime value. But it seems to be a lagging indicator.
How do we measure a leading indicator?
We think net promoter score (NPS) and customer satisfaction score (CSAT) could be a leading indicator.
How do we improve those?
We think average response time and average resolution time could be a contributing factor.
Can customer success directly influence those?
Data Sets Available
Are we measuring those metrics right now?
Yes. Manually. On Excel.
What do we need to better understand this function?
Well… it would be better if we could track a breakdown by
- Customer support representative
- Channel (e.g. email, text, website widget, social media private message)
- Type. of Problem (e.g. delivery delays, defective product)
- Peak time of responses
Are we collecting this data now?
Yes. But in 10 different locations.
Great. Better than nothing
Expertise on site
Do we have someone onsite to build these integrations?
Do we have someone onsite to run a regression analysis?
Is leadership willing to sponsor a new hire for this?
--End of Dialogue--
Once you start digging in, you will start realising that the underlying issue isn’t always the availability of the data.
Data set: Accessible & Clean
Even if you only have access to technology that may be a decade old, you should have a point-of-sale system that logs your sales transactions and merchandise information.
If you are an early adopter, you should have some of these in place
- Customer Relationship Management Software: Shopper
- In-store Sensors (Offline): In-store Traffic and Behaviour
- Inventory Management System: Product
- Google Analytics (Online): Website Traffic and Behaviour
Each one gives you a unique dataset which you can use for analytics.
There are two important points to highlight before moving forward. You have to ensure that the data is accessible and clean
Accessibility to Data
Regardless of the software vendor, you should have the ability to extract the raw files for your own work.
Take the point of sales system as an example, the data stored there might be required for accounting, inventory management and analytics.
In today’s cloud-enabled market, most software providers have and integration option or an open API that you can leverage to seamlessly integrate with the multiple tools that you will want to use.
If you have already committed to a vendor and they are not all that you thought they would be, change. Now.
Before it is too late. There are lots of cloud-based solutions that don’t cost more than $100 a month.
Cleanliness of Data
We have all heard phrases like “Data is the new oil”.
Oil is pretty useless unless it is refined.
An analytics platform is like a refinery that helps you make use of it. A refinery isn’t going to work as well if your data set isn’t as well maintained.
If the dates are all messed up or the products are mislabeled, the data is useless- clean it or trash it (depending on the state).
Word of Caution: Legacy vendors might not have kept up with technology best practices. Not all have proper database structures.
This will likely mean some shoddy patchwork to get a system up and running which can lead to complications down the line.
Difference between executive, managerial and operational dashboards
Managers figure this out pretty quickly.
They realise that the metrics they track,
...the metric that their leadership track and
the metrics they set for their teams are totally different.
All the metrics have to line up so that the entire company goes towards the same direction.
But they are rarely the look the same.
Just because a plane, a ship and a car are all going North, doesn’t mean they all care about the same thing.
Someone managing a 10,000 person organisation cannot be constantly checking into monitor the sales productivity of an individual sales rep.
Their dashboards look completely different.
Why? Because they serve a different purpose.
Purpose of Dashboards
The purpose of analytics dashboards is to give you a quick overview of the performance of your function role.
Within <2 seconds at looking at the numbers, you should have a good idea as to how you are holding up.
The next 3 seconds will be spent zooming into the numbers that are in the red and the next 5 minutes will be dedicated to figuring out why.
The Executive Dashboard
These dashboards tend to focus on a much longer horizon (typically a quarter to a year) of the business health.
The main focus will be around the financial and operational performance of the entire business rather than just one store.
During meetings, some may spend time diving into the performance of specific regions but one cannot expect executives to monitor the performance of one out of a chain of 500 stores every single day.
The Managerial Dashboard
These dashboards tend to focus on a mid-term horizon (typically a month to a quarter) of the business health.
The main focus will be around the operational performance of a particular business unit or a functional role.
Some examples include:
- Store sales performance
- Campaign return on investment
- Average delivery response times
The Operational Dashboard
These dashboards tend to focus on a relatively short-term horizon (typically a week to a quarter) of the business health.
The main focus will be around the operational performance of a functional role or initiative.
Depending on the role:
- B2B sales: Daily Productivity per rep
- Email marketing: Open and click-through rate
- Online Coupons: Redemption rate and Traffic Lift
These metrics and KPIs have to eventually lead back to the core pillars of the business (sales, product and shoppers) but they tend to be more focused around the results of initiatives.
Move onto the next section to view the metrics that matter.
203 metrics for retailers and brands to track
In this post, we will dive into each metric, discuss how to measure it and what it represents.
The importance of each metric differs depending on your company stage so don’t worry if you haven’t been tracking it.
Pick what makes sense for your stage.
To make this easier, we’ve split the metrics into 18 different domains so you can send this out to other team members.
1. General Accounting and Financial Analysis
We will be skipping over the purpose of accounting and focusing on the business goals that drive the underlying rationale behind some of the accounting and financial analysis.
- Cash in Bank:
Amount of money you have tucked away to ensure that you have enough runway to execute on your short to medium term strategies.
This includes hiring new staff, purchasing new software or increasing inventory levels.
- Net Income:
Formula: Sales revenue - All cost.
This is the amount of money you can expect to hit your bank account after different time periods.
- Operating Income:
Formula: Sales Revenue - Operating Expenses
Money leftover after you’ve deducted the expenses required to keep your business running day to day.
- Gross Profit:
Formula: Sales Revenue - Cost of Goods Sold
Amount left after deducting the direct costs associated with providing the goods or services.
- Sales Revenues:
How much your customer pays you for the goods or service.
- Other Revenues:
You organised a carnival to sell lemonade drinks
Carnival fees will be considered “Other Revenues”
- Cost of Goods Sold:
The direct expense relating to producing the good or service.
Item A cost $5. You sell it for $10.
The COGS is $5
At times, some forms of shrinkage may be included here to keep the books simple.
- Operating Expenses:
Includes Cost of Goods, Labour, and other day to day expenses (water/ electricity)
- Non-Operating Expenses:
Includes Depreciation, Amortization, Interest rates.
Basically things that you have to pay for but don’t seem to directly contribute to the production or service of your main line of business
- Gross Profit Margin:
Formula: 1- (Cost of Goods Sold/ Sales Revenue)
Ratio to represent how much of incoming cash will be locked up in providing the good or service
- Operating Profit Margin:
Formula: Operating Income/ Revenue
Similar purpose as gross profit margin but this captures the other overheads associated with running the business
- Cash Conversion Cycle:
Formula: Days of inventory outstanding -Days of sales outstanding + Days payable outstanding
This is a long formula that simply represents how long it takes you to turn cash into product and back to more cash again.
Formula: Inventory on paper- Actual inventory
Any loss inventory due to fraud or damage
- Estimated Market Share:
Formula: Current Sales / Total Addressable Market
Due to the lack of accurate data, most of these are simple estimations. Out of the total addressable market of the people who will and can buy your product or service. How many already have and will continue to do so.
- Growth Rate:
Formula: (Sales at End of Period/ Sales at Start of Period) - 1
Represents the speed at which your organisation is growing.
2. Sales Analysis
These metrics are the ones that most of the business leaders will be focused on but it is important to dissect these into relevant segments.
- Sales by Channel:
How much sales is each sales channel bringing in?
- Sales by Supplier:
Based on the merchandise supplied by different suppliers, how much sales did that generate?
Represents the financial reliance to individual suppliers.
- Sales to Date:
Represents the cumulative sales from a certain start date to this point.
Commonly used for monitoring the sales progress within a financial period (e.g. a quarter or a month)
- Average Order Value:
Formula: Total Sales / Total No. of Orders
Represents the average amount of money a customer spends per transaction. Depending on the industry, there can be quite a fair bit of fluctuations in these numbers. This represents a huge opportunity to drive these numbers up.
- Average No. of Products Per Transaction:
Formula: Total Quantity of Products Sold / Total No. of Orders
Represents the average number of products a customer purchased per transaction.
- Average Margin:
Formula: Cost of Product / Retail Price
How much is actually earned for every $1 sold.
- Sales Forecast:
NOTE: There are several types of forecast for differing purposes.
For a purely financial forecast, an estimate can be made using future marketing budgets and historical demands as a good estimate.
- No. of Returns:
Total number of exchange request
- No. of Refunds:
Total number of refund request
- No. of Exchanges:
Total number of exchange request
- Breakdown of Failed Sales:
Breakdown the above three based on
3. Associated cost
Represents what should be tackled first
3. Inventory & Warehousing
- Inventory Levels:
Per product, how many units are there store in warehouse (or the back of the store)
- Inventory Cost:
Formula: Ordering Cost + Carrying Cost + Cost of Deterioration
Represents the cost of managing inventory. Usually a significant percentage of the overall cost structure.
- Inventory Forecast:
NOTE: There are several types of forecast for differing purposes.
For an inventory forecast, a concrete number wouldn’t do much good. You will need a confidence interval.
With a 95% confidence an inventory level of 100 will meet the demand.
- Inventory Accuracy Rate:
Formula: Inventory Level Based on Physical Count / Inventory Level Based on Records X 100%
Some margin of error should be allowed but you should be alarmed if it is ridiculously off (e.g >7.5%)
- Days of Supply:
Formula: Inventory levels / Daily Demand
Number of days you can continue to supply the demand without going out of stock
- Holding Cost:
Cost associated with keeping inventory in the warehouse.
Sometimes opportunity cost is included.
- Safety Stock Percentage:
Percentage of extra stock kept just in case of a demand surge or a delay in supply.
- Safety Stock Cost:
Cost associated with keeping this extra inventory in the warehouse.
- Stock Turnover:
Cost of goods sold / Average inventory
Represents the number of times inventory is sold over a given period. (Higher the better)
- Sell-Through Rate:
Amount received from supplier - Amount sold to customers
Represents the wastage leftover of a particular product line.
- Receiving Accuracy Rate:
Accuracy rate between what receiver records and what was actually delivered.
- Cost Per Receive:
Cost associated with each received (more relevant to large warehouses and deliveries)
- Picking Accuracy Rate:
Correct Product Picked / Total No. of Products
A proxy as to how accurate warehouse employees/ robots are at picking and packing the correct merchandise.
- Lead Time by Supplier:
Time taken for a supplier to send the stock to the proper location from the time that the order was placed.
- Breakdown by Location:
Inventory stats broken down by location to see which to optimise first
4. E-Commerce Sites (On-site metrics)
Most of these metrics can be obtained using a free tool- Google Analytics.
- Total Traffic:
Sum total of all traffic to the site
- Traffic Source:
Source of traffic to the site (e.g. Facebook, Instagram, Affiliated Sites)
- Landing Page:
Page that people land on
Proxy as to which pages rank more highly on search engines or is the most linked in by other sites.
- Bounce Rate:
Percentage of visitors who leave the site without any activity on site.
Note: If this number increases too much it might hurt your search ranking because it is a signal that users didn’t find what they were searching for.
- Page Views per session:
Represents the number of pages visited by visitors per session.
- Average Time Per Session:
Formula: (Total time spent on site from ALL sessions)/ Total number of sessions
There might be some correlation between the increase in average time spent per page but you shouldn’t depend on this metric alone because an increase in average time spent could be due to poor UX that is taking users longer to perform the actions they require.
- In-Site Search:
If you have a search bar in your site, there will be 3 metrics to be optimised:
1. Time after search
2. % Search Exits
3. % Search Refinements
These represent the search intent of users who have already entered your site and are looking for something else.
- Email Capture Rate:
Percentage conversion of visitors who leave their email
Pro tip: Remember to Segment
Based on form factor (Email form v Pop-up)
Based on offer (Exclusive deals v $5 off first transaction)
Based on pages (Main page v Blogs)
- Peak Traffic Times
Depending on your technical architecture of your site, you might have to scale your servers based on peak traffic times.
If you are running A/ B tests, these traffic times will be when most of the data from your experiments get collected.
- Device type breakdown
Revenue breakdown based on the device type (Desktop, Mobile, Tablet)
Focus on optimising responsiveness across screen sizes
- Sales Per Country (State)
Revenue breakdown based on the I.P. address of the user’s device
If you were using a website builder (Shopify or Woocommerce), using their analytics based on the shipping location may be more accurate.
- Conversion Rates
Formula: (No. of purchasers / No. of site visitors)
Note: The formula above is for the macro conversion but there should be other optimisations between other pages (E.g. Product page to Add to Cart page) as well.
- Cart Abandonments
Percentage of visitors who have something in their cart but left the site before purchasing.
Note: Try tracking this metric across longer time horizons because there will be instances where shoppers return the next day.
- Credit Card Errors
Percentage of “purchasers” who have encounters at check out due to credit card errors.
- Product: Views
No. of times this product SKU was viewed
- Product: Add to Cart
No. of times this product SKU was added to cart
- Product: Conversion Rate
Formula: (View)/(Purchased) per product
- Multi-Channel Funnel:
MCF, as known as attribution modeling, attempts to measure the impact of each phase of the shoppers’ journey.
E.g. Website visit -> Ad on Google -> Ad on Facebook before a purchase
5. Physical Stores
- Sales per location:
Sum of total sales over a certain time period per location
- Sales per salesperson:
Sum of total sales over a certain time period per sales person
- Sales per square foot:
Sum of total sales over a certain time period per square foot
- Cost per square foot:
Cost of rent & utilities over a certain time period per square foot
- In-store Traffic:
Total foot traffic that enter your store
- Area Foot Traffic:
Foot traffic around the area of your store
- Area to Store Conversion:
Formula: (In-store Traffic / Area Foot Traffic)
Represents the conversion percentage of people walking around outside your store vs those who actually enter your store.
- In-store Heatmap:
A visual representation of which parts of the store has more foot traffic based on number of visitors and time spent
- In-store Hotspots:
A visual representation of which parts of the store has more foot traffic based on time spent
- Time Spent In-store:
Formula: (Total time spent in store from ALL sessions)/ Total number of in-store visits)
- Peak Periods:
Peak periods based on in-store traffic.
Should be used for manpower planning so that conversion rates don’t suffer
- Average Queue Length:
Average time in queue and the number of people waiting in line.
- In-store Conversion Rates:
Formula: (No. of completed Transactions/ In-Store Traffic)
- Revenue per Visitor:
Formula: (Revenue / Total no. of visitors)
- X- Day New Product Sales:
For new products, what was the sales for the first
- 3 days
- 7 days
- 30 days
- Estimated Sales Margin Range:
Before pricing the product, an estimation should be made based on historical sales and competitive data to estimate the margin of a product.
- Sales by Brand:
Sales revenue broken down by brand
- Sales by Category:
Sales revenue broken down by categories
- Sales by Supplier:
Sales revenue broken down by suppliers
- Sales by Product Attribute:
Sales revenue broken down by product attributes. Some examples that merchandisers take into account
- Packaging and Unit of measure
- Product Lifecycle:
Estimated amount of time before a product “expires” and becomes “obsolete”
- Fashion: When the next season starts
- Grocery: Based on the label on the products
- Quantity of Product Reviews:
No. of reviews for a specific product
- Quality of Product Reviews:
Average Quality rating (E.g. 1-5) for a product
- Supplier Reviews (Quality):
No. of reviews for a specific supplier
- Supplier Reviews (Quantity):
Average Quality rating (E.g. 1-5) for a supplier
- Seasonality periods per product:
Product attributions and seasonality of these attributes.
- Season 1: Black, Fur, Wet does well
- Season 2: White, Wool, Dry does well
- Product Complements:
Which products tend to be purchased together
- Product Substitutes:
Which products tend to be seen as competition to each other
- Sales per packaging:
Depending on the products, some products could be sold in different packages. For example;
- Coca Cola (1 can, 6 pack, 24 cans)
This can also be analysed as the same product.
- Recommendation Hit Rate:
Percentage of Total number of successful recommendations
- Up-sell Hit Rate:
Percentage of recommendations that were up-sells
Percentage of recommendations that were cross-sells
- Historical Pricing:
Shows the historical price points across time
- Sales per price:
Formula: Total revenue per historical sales price
- Elasticity Estimate:
Formula: (% Change in Quantity Demand) / (% Change in Price)
- Competitive Pricing:
Table of competitive pricing against yours and the estimated deviation
8. Digital Presence (Inbound, SEO & Other Unpaid)
- Site Traffic from Organic:
Total site traffic from organic traffic sources
- Blog Traffic:
Total site traffic to blog content
- Blog Conversion Rate:
Purchases from blog traffic/ Total site traffic to blog content
- Blog View Frequency:
How often your blog is visited and revisited
- Revenue per blog post:
Revenue driven from blogging/ No. of Blog Post
- Average Monthly Impression per content:
Total monthly impressions based on search, social, referral and others / Content
Represents the return on impressions for each new piece of content that is pushed out
- Common Search Queries:
Search terms that lead visitors to your page
- Page Rank on Search Engines:
Rank on search engines (broken down by search terms)
- Keywords Ranked:
Keywords that your site ranks for (broken down by page)
- Domain Authority:
Prediction of how well search engines will rank your site based on its relevance for a specific subject area
9. Email Marketing
- Email Open Rate:
Percentage of emails that get opened
- Email Click-Through Rate:
Based on the content of the emails, percentage of subscribers that click the links within the email
- No. Of Spam Complaints:
Total number of spam complains
- Unsubscribe Rate:
Percentage of subscribers that unsubscribe to your email
- Device Breakdown:
Ratio of Mobile : Desktop : Tablet : Others
Note: Based on where they first opened the email
- Revenue per subscriber:
Formula: (Revenue Generated from email) / (No. of subscribers)
- Revenue per email:
Formula: (Revenue Generated from email) / (No. of emails sent)
10. Social Media Engagement
- Social Media Fans:
No. of followers on each social media platform
- Social Media Engagements:
Note: There are different measures as to what constitutes an engagement depending on the platform
- Facebook: Like, Share, Comment
- Reddit: Like, Share, Comment
- Youtube: Thumbs up/ down, Share, Comment
- Post Type:
Total number of post based on type
- Image (Image with caption/ cute animal)
- Performance breakdown by post type:
Engagement breakdown based on image type
- Impression to Website Traffic:
Formula: (Website traffic from post) / (Impression per post)
- Post to Website Traffic:
Formula: (Website traffic from post) / (No. of post)
- Revenue per post:
Formula: (Sales driven from post) / (No. of post)
11. Digital Advertising
- Cost per impression:
Ad budget attributed to impressions / Total no. of impressions
- Cost per click:
Ad budget attributed to clicks /Total no. of impressions
- Customer acquisition cost:
Ad budget / Total no. of new customers
- Affiliate performance:
Affiliate-Driven Revenue/ Total no. of new customers
- Campaign performance:
Represents how well a particular advertising campaign performed (can be broken down into LOTS of other secondary dimensions)
- Keyword performance:
Represents revenue driven by a particular keyword
12. Promotional Analysis
- Number of Promotions:
Total number of promotions (Segment into active v inactive)
- Revenue Contribution:
Revenue from transactions that have a promotion applied
- Lift from promotions:
(Actual sales with promotion) - (Estimated Baseline without)
- Breakdown by promotional attributes:
Revenue breakdown based on promotional attributes
- 2 for X
- $5 off
- Buy one get one
Which of these are getting you more traffic?
- Breakdown by product:
Revenue breakdown based on product SKUs
- Breakdown by category:
Revenue breakdown based on categories
- Breakdown by brands:
Revenue breakdown based on brands
- Breakdown by channel:
Revenue breakdown based on sales channel
- Estimated Cannibalisation:
Opposite of Lift from promotions if the Lift is negative
13. Customer Support/ Success
- Average time to first response:
(Total time taken before first response)/ (No. of Request)
- Average resolution time:
(Total time taken to resolve all issues)/ (No. of Request)
- Total Sessions per day:
Include EVERYTHING that a customer support rep has to handle
- Web-chat sessions:
Total number of web-chat session and a breakdown of how long they take
- Social Media chat sessions:
Total number of social media-chat session and a breakdown of how long they take
- Email chat sessions:
Total number of email-chat session and a breakdown of how long they take
- Text- Sessions:
Total number of text-chat session and a breakdown of how long they take
- Overall Channel breakdown:
Breakdown of time taken on
- Social media
- Average number of opened tickets:
Opened tickets are issues that are currently worked on at this moment.
- Average number of resolved issues per day
Total number of issues per month / Days in the month
- Average backlog:
Backlog represents the number of unopened issues that the team has yet found the time to resolve
- Net Promoter Score per case type:
Breakdown the NPS based on the problem they faced
- Delivery delay
- Damaged product
- Net Promoter Score per Rep:
Breakdown the NPS based on the customer support representative
- Resolution Time per Rep:
Average time taken for a customer support representative to solve an issue
- Resolution Rate per Rep:
Percentage of issues that a customer support representative can solve
- Average Ramp up Time:
Average time taken to go from a complete noob to an average performer
- No. of session breakdown based on different purposes:
Breakdown the main reason why customers are contacting you
- Sessions resolution time breakdown based on purpose:
Based on the above, measure the time taken to resolve those issues.
- Resolution rate based on purpose:
Measure the resolution rate of those issues.
- Historical LTV:
(Total sales revenue to date) * (Average margin) per shopper
- Predictive LTV:
Based on churn rate, calculate estimated lifetime value
[Churn = (1/ LTV) ]
Lifetime (in months) * Average sale * Average Margin
- Cohort Analysis
Cohort analysis breakdowns the transactions that occurred within a month into new and returning customers to see if the number of returning customers is steadily increasing.
- Lasagna Model
The next level for cohort analysis.
Breakdown the returning cohort into further pieces to see when they first came.
Example: This month is April 2019
- 30% of returning shoppers came from Mar 2019
- 25% of returning shoppers came from Feb 2019
- 25% of returning shoppers came from Jan 2019
- 20% of returning shoppers came from before 2019
- New v Returning Ratio
Ideally this ratio should skew towards returning shoppers over time so that you can shift customer acquisition efforts to customer retention.
- Return Rate against time
Percentage of shoppers who return after the first month, second month, third month…
The goal is to drive these numbers up.
- Time between Visits
Average time between visits broken down by visit number
- Between the 1st and 2nd: 90 days
- Between the 2nd and 3rd: 100 days
- Between the 3rd and 4th: 95 days
A decrease in these numbers represent a higher shopper engagement score.
- Referral Rates
Percentage of current customers who are willing to refer others
- Net Promoter Score
A measure of how likely a customer is willing to refer your brand to their friends
For a benchmark:
- Apple’s score is 72
- Nike’s score 32
- Customer Satisfaction Test
A measure of the level of satisfaction with a particular service.
CSAT measure satisfaction levels whilst NPS measures loyalty
- Customer Acquisition Cost
Total cost associated with acquiring a new customer. Examples of cost that can be included
- Sales commission
- Ad budget
- Cost to create creative
- Software involved
- Churn Rate
Percentage of shoppers who leave your service forever
- Loyalty tier breakdown
Breakdown number of shoppers based on their loyalty tiers
- No. of points in circulation
If you offer loyalty points as a discounting mechanism, you will want to track the number of points in circulation for accounting purposes.
- No. of Loyalist
No. of shoppers who fall under “loyalist”
These individuals have
- An above average NPS score
- Seem to never churn
- Will always offer their feedback
Retailers should always have some idea of what market segment they are going after and what can be considered “loyal” in that market.
- Total No. of Distinct Segments
Measure of how many shoppers segments you have based on their purchasing behaviour.
Not everyone shops the same way.
- Market Segment Growth
Based on your target shoppers, how fast is the market growing?
Example: If your store sells green tea, how fast is the green tea enthusiast market growing?
15. B2B Sales
- On Target Earning (OTE):
Amount that a sales rep should expect to take home (as thus how much they should be expected to sell)
- Average Sales per Rep:
Formula: (Total B2B Sales) / (No. of reps)
- Average Sales Cycles:
Represents the average time taken to close a new deal (from start to finish)
- Average Win Rate:
Formula: (No. of closed deals) / (No. of initial deals)
- Overall Funnel Conversion:
Conversion rate from:
- First contact
- Meeting / Email
- Invoice sent
- Payment made
- Average Deal Size:
Formula: (Total B2B Sales) / No. of Deals
- Capacity per rep:
Number of calls, emails and meetings a rep can take in a month
- Average Ramp Time:
Average time taken to go from a complete noob to an average performer
- Average Margin per rep:
Represents one variable in determining if reps are good enough to close deals without having to offer too large a discount
- Percentage Hitting Quota:
Overall percentage of reps that attain their monthly quota.
- You don’t want this at 100% because it means you guys aren’t stretching enough
- >80% should be hitting quota
16. Human Resources
- Revenue per Employee:
Formula: Total Revenue/ Total number of employees
- Average Time since Promotion:
Average time since an employee has been promoted
- Cost per New Hire:
Cost (sometimes this could just be the time invested) to hire a new candidate
* Time spent interviewing failed candidates should also be included
- Performance Appraisal Rating:
Rating of performance from managers
- Peer Appraisal Rating:
Rating of performance from peers
- HR to Employee Ratio:
Ratio of HR professionals to Regular Employees
- Turnover Rate:
Average time spent within the organisation before a personal leaves
- Time to Hire:
Average time taken to hire for an unfulfilled position
- Ramp Time:
Average time taken to go from a complete noob to an average performer
17. Delivery and Fulfillment
- Average Fulfillment Time:
(Total time taken to fulfill all orders) / (No. of fulfilments)
- Overall Delivery Cost:
Total cost associated with delivering product
- Delivery Cost per Unit:
(Overall delivery cost) / (Total no. of units)
- Perfect Order Rate:
Percentage of orders that were delivered perfectly
- On time
- Right product
- No customer complaints
- No. of Returns:
Total number of returns
Remember to segment by:
- Customer segment
- Fulfillment centre
These attributes should help you diagnose the problem later
- Cost of Returns:
Total cost of returns
- No. of Exchanges:
Total number of exchanges
Remember to segment by:
- Customer segment
- Fulfillment centre
These attributes should help you diagnose the problem later
- Cost of Exchanges:
Total cost of exchanges
- No. of Backorders:
Total number of orders in the backlog left unfulfilled
- On-time Shipping Rate:
Percentage of orders that get fulfilled on time
- Lull Time for Trucks
Percentage of time that trucks are not in use
- Cycle Time:
Represents the average time taken to manufacture a single product. From start to finish.
- Equipment Productivity:
Represents the percentage uptime that the manufacturing equipment is in use
- Labour Productivity:
Represents the percentage uptime that labour is active
Represents the number of products that can be produced each cycle
9 metrics for start-ups and small businesses should focus on
For a small business, focus is the name of the game. You can’t do everything so you have to focus on only a few core metrics that really drive your business.
For most businesses, this should fall under three board buckets: Sales, Products and Shoppers.
These metrics should be tracked relentlessly and early warning systems should be going off if one of these isn’t going as planned.
Is money coming into the business as expected or are you losing some months and why?
Whether it is money locked up in stock investments or new hires, you have to have a rough estimate of the payback period for this cash.
2. Sales Revenue
How much are you making from monthly sales? This is a number that your whole executive and management team should be familiar with.
There should be no unexpected surprises that causes this number to fluctuate month on month.
Getting this metric right is enormously important because once you track this, you can start investing in the underly infrastructure to understand what moves it.
3. Net Income
After subtracting all the costs of running the business, how much do you actually have left? Knowing this number allows you to answer questions like
- Is the business healthy?
- Can we afford to invest in growth?
- How much can we afford to invest in long-term strategies?
- Are our previous investments paying off?
- Should the next quarter be focused on growing the top-line or managing cost down?
4. Average Margin
This is the marginal cost of doing business. If you were to sell the next product, how much would that make you?
You have to balance this against the volume of transactions that your business is currently doing and this should inform your overall business roadmap.
If you know that you plan to beat the competition in the lowest cost game, you have to make it up in volume. If you plan to win in terms of quality and service, you have to have a healthy margin to keep the service sustainable.
5. Average Order Value
How much is every customer spending with you every transaction? Is this number going up or down?
Steep decreases in the number could signal
- The entrance of competitors in your space that you’ve heard of because your loyal shoppers are purchasing goods somewhere else.
- A poor merchandising mix that’s confusing to shoppers and causing product cannibalisation
6. Sales by product
Breaking down the sales by products is a great way of deciding which other products to bring in and which products should be stopped because their sales just don’t justify the cost of bringing them to market.
Poor merchandising choices can not only hurt your sales in the short run but your entire brand in the long run.
7. Customer Acquisition Cost
Note: You should use this metric with the Lifetime Value (LTV) metric because this metric alone doesn’t tell you much about your business.
Customer acquisition cost represents how much it cost you to acquire the next customer. This should include any salary, advertisement dollars, marketing asset creation and any software purchased to help you acquire that next customer.
8. Churn Rate
Churn represents the percentage of shoppers from a specific cohort that have stopped returning to your store.
This is an early warning indicator of the health of your business because the longer a shopper is willing to stay with you, the less resources you have to commit to acquiring new customers.
9. Lifetime Value
Customer Lifetime Value represents valuable a customer is to you.
This represents the (a) amount of sales generated per customer per month * (b) average margin * (c) estimated number of months a shopper will spend with you.
If you can improve any of those levers, your business will start looking healthier instantly and you can more freely invest in acquiring customers knowing that they are likely to be worth much more than before.
Operational Analytics highlight
As data has become cheaper to store, analytics products seem to have proliferated the marketplace promising lots of new ways to visualise and understand ones’ data. Now, there seems to be another field in the data analytics space emerging. Operational Analytics.
What is Operational Analytics?
Operational analytics is the practice of leveraging data to augment or optimise functional roles or processes within the company.
How is this different from analytics in 2018?
Majority of the analytics has been heavily focused on managerial or executive decisions that aid in crafting strategy or company-wide initiatives but do little to influence the next small decision that needs to be made within the organisation.
A dashboard that focuses on the overall Net Promoter Score (NPS) of the company isn’t directly suited to helping customer success representatives decide how best to support their customers.
What are some examples in retail?
Haste is an operational analytics platform for retailers and brands so I will be focusing on 2 areas that I’m more familiar with.
Management’s main concern over merchandising tends to be around which products are still in the warehouse and how to get rid of them but these questions don’t always help merchandisers decide on the next product to bring in.
They are more interested in changing shopping patterns, hot new materials or the attributes of top performing products.
Management’s main concern around marketing and promotions tend to be around the outcome of the promotions like the revenue contribution and the lift from sales. Although important, one needs to step back and ask, “How best do I come up with a promotional plan”.
What sort of promotion should it be? What should the eligibility criteria be? How should the promotion be?
Another dashboard and analytics tool should be dedicated analyzing these sorts of questions.
Where to start?
Focus on the core pillars of your business first: sales, products and shoppers.
If you know that these 3 pillars are what drive your business, invest in fully understanding them before moving onto the gap between these things.
- Product to Shoppers: Advertising optimisation
- Shoppers to Sales: Conversion rate optimisation
- Sales to Products: Cash flow optimisation
Leveraging Analytics for Long Term Success
Analytics isn’t some new secret in today’s world.
How we use it, however, is going to be the main differentiator.
If you outlast or outsmart your competitors, you can still edge them out of the market. Those that started prematurely might not have fully formed a well-thought-out strategy, and that’s the arbitrage.
Are you going to take advantage of that?
Start with this thought framework
Don’t invest a single dollar in any analytics implementation until you can clearly answer these questions-
- What problems do I want to solve? (done)
- What sort datasets do I have on hand?
- What is the level of expertise of my team?
Always start with defining the problem and the optimal outcome. Here are some examples-
- We want to improve our cash flow.
- Get a basic analytics tool for detective work
- There is an opportunity to reduce our cash locked in inventory by 40% by the end of the year
- Get a more advanced tool to find out how
- Oh! If we improve our demand forecast models by 20%, we don’t have to carry that much inventory in our warehouse
- Hey, look at that. More money in the bank that can be reinvested into other assets to grow the business.
Yes, it may seem like an optimistic outcome, but you will be surprised how often these little things happen.
Then start enabling the rest of the organisation
Once you have a clear idea as to what you want to get out of analytics, it’s time to empower the rest of your team.
There are dozens of techniques to solve the same problem, and you have to pick a solution that best suits you.
Many make the mistake of falling into the technology hype and purchase a tool that doesn’t actually fit their specific workflow.
So how do you pick an analytics tool that your team will feel comfortable using?
Assuming you are not a subject matter expert, let them choose because they are more aware of
- Limitations of the data available
- Their own limitations
Before you had the tools in place, they were probably already using some other way to get the same outcome.
Whether it was complex SQL queries or super pivot tables in Excel, they know how to work with the data they have.
They are less likely to recommend a tool that is too complicated for them to use and they are more suited to screen out retail analytics solutions that cannot work with the data set.
Invest in enablement
It is always tempting to think that you can get by forever without purchasing an analytics implementation tool because your team can always find a better way out. Here are some facts;
- The data scientist or business analyst on your staff should probably be the most well paid
- They are likely to be under-appreciated because many managers don’t know how to set goals and objectives for them
- They will leave soon because their skills are in hot demand and the tools that you are offering them are taking too much of their time.
- We are not saying that you HAVE to purchase an analytics tool for them. You are going to need some financial justifications.
Now let’s think about it this way-
- You pay them $50,000 a year (on the lowest bound)
- The analytics tools reduces the average workload and time taken by only 2x
- You just reduced their workload to such a degree that they now…
- Can finally spend their time looking for other ways to improve your business
- Complete 10 more tasks in the same time frame
- They will love you
You just made your analytics operations 2x more efficient for the cost of software.
How does the software compare?
Does it cost the same as hiring another analyst?
Depending on the stage of your business, these assumptions might or might not hold true. If you haven’t even got the basic tools, getting one now might increase the efficiency by a factor of 10x.
Building a long-term data strategy
In essence, this strategy comes down to establishing the infrastructure required for analytics teams to move fast and execute. These teams need three things-
Depending on the contents of the raw data, some departments might be uncomfortable disclosing the numbers to the analysis.
Say, a marketing department wildly overspends on an unsuccessful promotional campaign. It might be embarrassing for them to share these numbers with others in the organisation.
Your duty as someone part of this organisation (even if you don’t have a leadership title) is to encourage this sharing of data.
Like it or not, the success of this strategy hinges on the involvement of the senior management.
It doesn’t matter how much you invest in your analytics team if their work is not going to be recognised by another department.
They are going to get jaded and just work on improving some other part of the organisation while that portion remains as it is.
Individuals who possess strong analytical skills are in hot demand.
They are generally hard to recruit and retain.
Don’t make them work with old tools that take forever to load.
Try letting them choose the tools that they prefer working with because they will be more aware of the limitations of the various tools and the data set available.