How I Cut Cost Per Lead by Half at an EdTech Company

By
Abhilash Jose
I’m Abhilash Jose, a digital marketer with a background in data analytics and marketing operations. I combine SEO, digital marketing, and data to understand what works,...
Disclosure: This website may contain affiliate links, which means I may earn a commission if you click on the link and make a purchase. I only recommend products or services that I personally use and believe will add value to my readers. Your support is appreciated!
Marketing Analytics · Case Study

Picture two campaigns running side by side. One brings in 500 leads a month, the other only 300. At first glance, the answer seems obvious, until you look at what each lead actually cost.

That question shaped how I worked across offline campaigns and locations in Kerala, working closely with the sales team to compare how leads moved through both funnels, ran ongoing trials on where and how leads were generated, and used that data to guide budget decisions and streamline reporting.

Scope: offline campaigns, Kerala region · Reviewed on a recurring cycle

  • 56%
    lower Cost Per Lead
  • 60%
    less time spent on reporting

The Problem

Lead generation was running across several campaigns and locations, but performance wasn’t consistent. Total lead count was the number everyone looked at first, and it was misleading. A campaign could bring in a large volume of leads while quietly costing far more per lead than a smaller, better-targeted one, or generating leads that rarely converted once they reached sales.

Without a shared view across both teams, budget and manpower decisions were being made on partial information. On top of that, the KPI reports used to track performance were rebuilt from scratch every cycle, leaving less time for the analysis that actually mattered.

My Approach

CPL alone couldn’t explain why one area outperformed another, a cheap lead that never converted wasn’t actually efficient. Understanding that meant looking at two funnels together: the one marketing controlled, and the one sales controlled.

The marketing funnel: from asset to lead

Offline leads started with a physical presence at a location (“asset”), for example, malls, supermarkets, hypermarkets, airports, railway stations, small stores, or clothing stores. Each asset had its own cost and lead potential:

  1. Identify relevant assets
  2. Assess each asset’s potential and cost
  3. Decide BDE (Business Development Executive) deployment, sized to the asset’s cost and the lead volume it needed to justify
  4. BDEs pitch to relevant customers at the asset
  5. Leads are collected

The sales funnel: from lead to order

Once marketing handed a lead over, a separate funnel decided whether it turned into revenue:

  1. Lead allocated to a sales rep
  2. Lead dialled
  3. Lead connected
  4. Demo booked
  5. Demo conducted
  6. Order placed
MARKETING Assets identified Cost & potential assessed BDEs deployed Customers pitched Leads collected handed to sales SALES Leads allocated Calls dialled Calls connected Demo booked Demo conducted Order placed
Marketing and sales funnels tracked together, a lead’s cost only matters in light of where it dropped off, or didn’t.

Working with both teams, I compared performance across both funnels, not just which asset produced the most leads, but which ones produced leads that actually survived the sales funnel through to an order. Some sources drove strong lead volume with a weaker path to a demo; others produced fewer leads but converted noticeably better. Looking at CPL on its own, without this second half of the picture, would have pointed budget in the wrong direction.

Ongoing testing, not a one-time fix

This wasn’t a single analysis done once. New assets and approaches were tried on a rolling basis, given a fair trial period, and reviewed against consistent funnel data rather than a single month’s numbers or gut feel. Some that looked promising early didn’t hold up once measured against full-funnel conversion and were phased out; others were given more support and scaled up once the data backed them. The process itself, test, measure, decide, adjust, repeated continuously as conditions changed.

Test Measure Decide Adjust
The review cycle: nothing was scaled or stopped based on a single data point.

Based on this recurring view across both funnels, budget and BDE deployment were shifted toward the sources that were both cost-efficient and converting through to orders, while support was pulled back from ones that consistently underperformed across full trial cycles. In parallel, I rebuilt the KPI reporting sheet into a standardized, semi-automated format covering both funnels, so the same view could be refreshed each cycle without being reassembled by hand.

A cheap lead that never becomes a demo isn’t efficient, it just moved the cost from marketing’s sheet to sales’.

The Result

Reallocating budget and effort toward the better-performing sources, based on this recurring test-and-review process, brought Cost Per Lead down by 56%, the same overall marketing spend produced leads at less than half the previous cost.

Before After Higher CPL 56% lower
Cost Per Lead before and after data-led reallocation.

The reporting rebuild had its own payoff: what used to take hours of manual sheet-building each cycle came down by about 60%, freeing that time for actually interpreting the numbers rather than compiling them.

  • 11
    funnel stages tracked across marketing and sales
  • 20+
    asset categories evaluated on a rolling basis
  • 2 teams
    marketing and sales aligned on one shared view

What I Learned

This was the project that pushed me from doing marketing to analyzing it. I started with Excel and Google Sheets simply to make sense of campaign performance, and that need kept growing, eventually pulling me into SQL, Power BI, Python, and statistics to work with data at a deeper level.

It also changed how I think about marketing work generally: execution and analysis aren’t two separate stages, and a decision is only as good as the number of review cycles behind it. The work gets done, the results get measured, the data gets read, and that reading is what decides what happens next.

Tools Used

  • Excel
  • Google Sheets

Concepts Applied

  • Cost Per Lead (CPL) Optimization
  • Multi-Funnel Attribution
  • Sales & Marketing Alignment
  • Rolling Experimentation
  • KPI Dashboard Automation
Share This Article
Follow:
I’m Abhilash Jose, a digital marketer with a background in data analytics and marketing operations. I combine SEO, digital marketing, and data to understand what works, why it works, and how to improve it.I’ve worked on marketing campaigns, funnel analysis, reporting, and performance optimization, helping businesses make better decisions using real data.Today, I’m focused on helping businesses get found online, attract the right audience, and turn traffic into measurable results through practical digital marketing strategies.Learn. Test. Measure. Improve.