4. Optimizing Your Funnel with Data Slicing and Analysis

  • Sameer Mehta Ex-SVP Growth & Marketing, WealthDesk (acquired by PhonePe)

When you look at, let’s say, a Facebook ad report, it would just give you the basic data — you ran an ad for so many days, this was the number of impressions, the reach, clicks, cost per click, how many leads, how many bookings, and what was spent.

What I’ve seen is that a lot of people don’t go deep enough into the data to apply different layers. What you actually need to be doing over here is really trying to understand the funnel your user is going through. This is important because it will give you input on how well your ads are performing across the funnel.

If you're able to add layers over here, and you're able to add some commentary, then this will give you very rich data. You can refer back to your data to check on previous experiments and the learnings from those.

You can do this at a small scale also.

For example, when you create WhatsApp content and you send it out, you have to see how many people are responding to it, you have to see how many people are clicking on any links that you added to your message. Did the clicks spike on a particular day?

Slide titled ‘Apply more layers to your funnel,’ describing a data table of ad performance metrics across multiple campaigns, including impressions, clicks, CTR, cost per click, leads, and conversions, highlighting granular performance analysis
Sameer’s slide on showcasing layers added to ad analytics.

You have to apply more and more layers to every funnel that you use, and you have to apply these layers stage wise.

For example, at an awareness level, am I seeing something that is working better? Are some keywords resonating better than the others. Refer to the on demand’ versus daily cleaning’ example from earlier.

Also, if you're doing this pan India, then it’s very important for you to see how your data is looking across different cohorts. You will get very useful insights from slicing your data for tier one city versus tier two/tier three etc.

You might look at creating content in English vs regional languages. You might want to create content in Gujarati if you are a financial services company, because that’s a very active state for financial services.

Another thing to be wary of is that a lot of companies end their analysis of the funnel at the cost per lead level or what is my CAC”. While at the aggregate level, you will look at CAC and LTV and figure out how it’s working.

Slide titled ‘Layers,’ describing different dimensions for analyzing funnel data, including stage-wise metrics, customer tiers, language segmentation, device type, and revenue, highlighting deeper analytical slicing for optimization
Sameer’s slide on “layers” you can add to your funnel data.

If your tracking methodology is good, you will also be able to figure out that a particular channel is giving you better, higher paying users. Or that some ads are actually getting you users that are retaining for longer, or signing up for longer subscriptions.

That's why it is very important that you take this analysis all the way down to your revenue and retention metrics. This is very rich data, which you utilize to figure out how users are engaging with you and your product.

I’ve seen most companies not doing it, because typically they only focus on the total number of leads, or just top line growth. They don’t necessarily look at it from an ROI perspective, or what is my bottom line.