• goal setting
  • metrics
  • Digital Marketing
  • Podcast

#InConversation: Understanding Metrics and Goal Setting

Read More
All Posts

#InConversation: Understanding Metrics and Goal Setting

Defining metrics and goal setting

Once every week, I rope in an an industry lead to talk at length about the latest trends in digital media. In this conversation with Ankit Agarwal, we talk about how media planning and buying is evolving with the rise of mobile.

For today’s topic, we would want to understand that how do marketers identify and go about defining metrics for themselves?  

Infact, before we dive deep into e-commerce and goal setting, what is important is to just lay down once the metrics which are available for a specific marketer. When they start their paid market in a paid campaign, so what sort of metrics are available in the first place or what sort of metrics should a marketer definitely configure in their dashboard while they kick off any campaign, be it paid or organic for that matter?

In the context of paid media, there are two key metrics that can be looked at from a 30,000 feet perspective. The first one being, Cost per Click (CPC) or the Cost per Impression (CPM) depending upon what the media buy type is. As you dive a bit deeper and marry campaign data to web analytics data, the metric changes to Click to Landing Ratio, Bounce Rate and Time spent on page. This emerges from the fact that not all clicks will result in actual to visits on the page and not every visitor will stay on the page (bounce).

defining metrics for marketeers

 

While these are vanity metrics, these help you evaluate your paid media sources and the quality of audience they have and what they are pushing to you. From a marketer’s perspective, if you are paying to acquire traffic, you definitely want the incoming traffic to engage with you and not just bounce off immediately.

Here are some benchmarks worth chasing:

On the last mile journey or bottom of the funnel, from a macro-conversion perspective, the metric to track closely is conversions ( sales, form submissions, registrations, payments). For instance, in the E-Commerce space, this could be as basic as Number of Purchases. As you grow and become more sophisticated, you would want to capture the Product Id / SKU or Product Category, Cart value and revenue generated.

From a regular executive dashboard, execs should track  

  • No. of Visitors ( Including Organic )
  • Bounce Rate
  • No. of Users purchasing
  • Number of Products/SKUs Sold
  • Total Revenue

Metric such as Product Sold, specifically helps you understand your targeting. For instance, did the user buy a pair of socks or did they buy a fridge? Because not only the revenue is different, but it also gives you an understanding of your campaign targeting and the end output.

For eg. If you are advertising on a fridge keyword are they actually buying a fridge or users come from a fridge buying keyword but end up buying socks.

What Next?

The sophistication can go beyond and then you could essentially start tracking LifeTime Value (LTV) of a customer. The challenge with LTV is that it’s not something your web analytics tool or your ad platform will give you. LTV needs to developed basis your audience, their shopping pattern and transactional behaviour. Day by Day, marketers are aggressively looking /analysing data and asking, specially in the e-commerce space. Key questions that are put on the table are:

  • Are the current acquisition efforts / channels scalable
  • Is the traffic acquired repeating their organic visits and transactions over a period of time ( Life Time)

The prophecy is simple-

  • All efforts should be aligned to drive users acquired via paid channels turn into loyal users such that they convert, transact organically. A portion your paid users must stick around and contribute to your sales and not costs.

Starting Point for E-Commerce Marketers ?

There is a saying that there should only be one tool in all. From an ecommerce standpoint at the end of the day, the campaign metrics has to sync in very closely with your business metrics which is Gross Merchandise Value ( GMV). There are more than dozen metrics in the funnel but the ultimate metric that really matters is the revenue that is generated or essentially Return On Investment (ROI) or Return On Advertising Spend (ROAS).

ROAS essentially says I spent a dollar and I got two dollars out. So I have a ROI of 1 or 100 per cent. On a very simplistic level, the end goals for an ecommerce based company would definitely be Return on Ad Spend or Return On Investment.

Beginners

For somebody who is just starting with user acquisition activities via paid media, Cost per Acquisition is a good metric to start with. This allows the marketer to start understanding how much you have to pay to acquire a customer for drive a transaction. Having said that, once the campaign has stabilized, this becomes a vanity metric. It is then that you essentially want to understand how much are you paying to get this customer and how much is the customer then paying you back by spending “x” amount of dollars. This ideally should get calculated over a period of time and not just from the first transaction. For an Ecommerce setup, return on ad spends or return on ROI is the metric at scale that should get tracked on.

Right time to think about LifeTime Value

Lifetime Value is an asset, and is a difficult metric to arrive and track in the early days. The key question is at what stage of journey in terms of scale, operations , etc should a marketer really define lifetime value of a customer?

Short answer - It depends. Long answer - There is no perfect time. Taking a step back, let’s first define the building blocks that marketers should be tracking on a daily basis:  

  • Average # of Repeat Orders
  • Average # of Repeat Sales
  • Average Frequency for repeat order

Now, there are two ways to approach this -

  • Order / Sales based Milestone

I have seen instances wherein at a scale of 2000+ daily conversions or 3000-4000 direct sales/day, is an appropriate time to evaluating LifeTime Value aggressively.

At this scale, one will start seeing statistically significant data, especially of repeat orders.

  • AdSpend Based Milestone

An alternate approach way would be, basis the amount you are spending. For instance, if you spend 20$ to acquire one customer and getting you are getting 3000 customers from paid channel, vary so if you do the mark it is eventually of a good amount 60,000$ that I am spending everyday to acquire those customers. That is a good amount of spend wherein I should really get serious about LTV or Lifetime Value.

Evaluating “Lifetime”

“I worked very closely with a Fashion ETailer in Toronto, which primarily sold Winter Jackets along with winter accessories. For a retailer like that you can’t even do a 30 day or 90 days LTV because they are going to buy one jacket and probably come back only next year because that's how the jackets are like 2000$ jackets. So for an e-commerce player like that, I think the focus was saying 12 months in the given of time so do we see them again in the same period. For eg.if they bought it in October today do they at least come, start exploring our site next october or next September. For them, it was kept at least 12 months.”

However, this would change. For instance, in case of hyper local grocery delivery business, I will definitely have a lifetime value, I would define it, let’s say, over a 90 day period and see did they actually like from me, how times they like it say 3 times or  5 times because see this is a business in which I should be buying every week. So that is why there is no number to be put in, which can apply for every field. It’s not very difficult to estimate based on what your business is,  like the few examples we discussed. The hyper local example is pretty interesting, so I would really think my lifetime value issues ideally be within like 30 day period if I am really doing it right because I should at least see the customer return in the next 30 days because if they are buying, say, vegetables or  fruits or whatever that ideally they should require within that 30 day period at least once and if they come buy from me, I have a very good understanding of saying that this customer comes back within a 30 days period so I should able to pay more to acquire them.

There is no fixed formula clearly. It obviously is a function of - the product you are selling, seasonality, your overall product catalogue, the repeat need of the product the, frequency and the recency model of course. Fundamentally the business should look at the data, the purchase data for a specific period of time, for a specific sort of users and see at what times do they repeat transaction,start chasing out for a specific user and take that as the lifetime itself and some of the transaction as a lifetime value.  There are those complicated formulas that exist on  internet in terms of actually calculating lifetime value but I think on a fundamental level the points that you’ve mentioned, those are the one’s you eventually start calculating. Lifetime value to start with might not be absolutely scientific and sacrosanct but I think your existing data should be able to tell you enough things. Lifetime value is a moving target as well and ideally, marketers should re-look at data and evaluate it, say every 60 days

Understanding Negative Return on Ad Spend?

Here is how marketers with a defined value of Lifetime approach customer acquisition.

For instance, if an ecommerce marketer spends 100$ to acquire a new customer. And the cart value is 80$. With a lifetime of 100 days, the average customer is expected to come back in 20 days. Average repeat transaction is worth 100$. With that as a premise, cost per acquisition becomes meaningful even if on the first transaction it might be negative but I think that's where it becomes interesting to look to a CPA from a cost per lifetime value instead of cost per acquisition.

What is extremely important to note is that calculation of Lifetime Value and Lifetime itself is an iterative exercise. It should be calculated at regular interval.

Loyalty in the Indian Context is a delusive idea. Should marketers in India follow this approach ?

While it is ok to have a negative ROI on the first transaction but it is important to approach it  from lifetime value perspective. Having said that, it is difficult to calculate the lifetime value in an atmosphere skewed towards new customers acquisition, app installs, either via app or via SEM or via display campaign, and coupling it with heavy discounting. Fundamentally that does throw the complete lifetime value construct of the grid. In the current scenario, the lifetime value is being pegged over the fact that may be in the next five years everybody is going to buy online and that's where the lifetime value comes in, the brand value comes in.

In reality, I don’t have the luxury of sitting with a brand and saying you know what accept the higher CPA  and we will try bringing it down over a period of five years.

Lifetime could be, say 2 years, it could also be 180 days. This is where business and the marketing need to work together and instead of having those idealistic dreams about what the lifetime should be, it should be more realistic and accepting that, yes we are an e-commerce company we want the customer to come back and buy from us in 30 days, so we are going to evaluate lifetime value, say, in a 90 days period.  Especially in the Indian market you see there is a lot of competition in e-commerce. I personally don't think there is a brand loyalty play has come in. I’m sure it is building up or will eventually go to a point but at this point, I would be looking more realistic and not look at those intangibles like brand loyalty. I still have a hard time in putting them in an excel sheet and measuring it. 

Looking for Red Flags

Some critical points to look out for:

  • Data Hygiene:  A lot of brands struggle with this issue where data is not tagged properly. For instance, search traffic is not getting tracked as a direct revenue and it is compared against direct traffic. Or for instance, social traffic is not tracked properly and all that conversion goes to some other channel. If you don't have clean data than you essentially going to make decisions that will be skewed. It is in these cases it is important to have external benchmarking done. For eg. Organic should be generally convert @ 4%. If I am seeing it at 40%, it is a great story yes but that's a red flag because then I want to understand why is that? Is that my paid traffic that is not showing or my organic traffic or something of that sort. So, measurement or bad measurement for that matter is a big red flag for me when I am looking at campaign metrics.
  • Calibrating CPA with Product Delivery: As a marketer, you need to factor in all the pitfalls that are possible post the transaction. So from CPA perspective I could be looking at making increment to CPA gains, with the ROI from the campaign which is improving day over day but the biggest challenge, is order cancellations. With a 20-30 % return and if you are not factoring that data into your campaign metrics, you are doing injustice to the business. A 100 rs CPA or a cost per sale and driving ton of sales with a 30 % return rate, actually translates to a much higher CPA. This is pretty much a function of saying either traffic is not qualified or traffic is being sent to the right product.
  • Quality vs Quantity: What is important 1 Mn visits with average time spend of 4 secs of 1000 visits with average time spend of 5 mins ? This is important to have quality of traffic monitored very closely.  For eg. if you look at a lot of the affiliate networks - most of the traffic sources are not clear. As a marketer who wants to focus on a real end goal, one should not fall for buying cheap traffic

While these are vanity metrics, these help you evaluate your paid media sources and the quality of audience they have and what they are pushing to you. From a marketer’s perspective, if you are paying to acquire traffic, you definitely want the incoming traffic to engage with you and not just bounce off immediately.

Here are some benchmarks worth chasing:

On the last mile journey or bottom of the funnel, from a macro-conversion perspective, the metric to track closely is conversions ( sales, form submissions, registrations, payments). For instance, in the E-Commerce space, this could be as basic as Number of Purchases. As you grow and become more sophisticated, you would want to capture the Product Id / SKU or Product Category, Cart value and revenue generated.

From a regular executive dashboard, execs should track  

  • No. of Visitors ( Including Organic )
  • Bounce Rate
  • No. of Users purchasing
  • Number of Products/SKUs Sold
  • Total Revenue

Metric such as Product Sold, specifically helps you understand your targeting. For instance, did the user buy a pair of socks or did they buy a fridge? Because not only the revenue is different, but it also gives you an understanding of your campaign targeting and the end output.

For eg. If you are advertising on a fridge keyword are they actually buying a fridge or users come from a fridge buying keyword but end up buying socks.  As well as Understand your target audience deeper for your keyword is very important.

vivek
vivek
Founder at Datability Solutions. Second Venture. Writes about technology, entrepreneurship and mobile. Avid cyclist.

Related Posts

Segmentation 101 - Moving Beyond Flat Personas To Retarget Better

In this age of personalized marketing, turning a blind eye towards seg[...]

Now Live: Push Notifications on Opera and Edge

Opera and Edge together constitute to about 5% of browser share global[...]

Now Live: A Fresh Way to Work With Your Subscription Prompt

Web push notification is a permission-based medium and you can use Sub[...]