The Science of Conversion Tracking: A No-Fuss Guide to Conversion Metrics
This guide provides an overview of the most common conversion metrics used in conversion analysis and explains how to measure and calculate:
- Conversion rate
- Conversion volume
- Conversion time
- Cost per conversion
Each subsequent post will be dedicated to providing you with an easy, step-by-step guide explaining how to calculate each of these four conversion point measurements, while also providing examples and pointing out key considerations to think about along the way.
I will also share how to determine the appropriate time horizon for a given conversion measurement, as that can play a critical role in the accuracy and reliability of a conversion measurement.
But before we dive in, let’s first establish what we’re talking about when we talk about conversion in the context of customer acquisition.
What is Customer Conversion?
Conversion is the process of a prospective buyer knowingly or unknowingly taking action that moves them from (or at least closer to) one stage in the buyer journey to the next. A conversion point is the transactional moment whereby this action occurs.
For example, take the email marketing aspect of a company’s go-to-market strategy. One conversion point they may track is the email call to action, such as clicking on a link to a report you want to share with them.
It is important to note that there are generally multiple conversion points during any specific stage in your customer’s buying process or buyer’s journey. For example, signing up for a newsletter or following your company on Twitter could both be examples of conversion points during the early awareness stage of the buying process.
How to Calculate Conversion Rate
What is Conversion Rate?
Conversion rate is a measurement of how effective marketing and/or sales have been at getting prospects to successfully move past a conversion point. It can also serve as a good measurement of the quality of traffic from a specific marketing channel. However, what it does not give you is a sense of actual volume to expect via a specific conversion point.
Conversion rate is calculated as the ratio of prospects to completed events/transactions during a particular time period or specified length of time (stay tuned for more details on the time element in a subsequent post).
Conversion Rate = (Completed Events / Total Transaction) * 100 (Time is Constant)
Conversion Rate = (Conversion Volume / Total Volume) * 100 (Time is Constant)
Example Business Application of Conversion Rate
For example, if we were looking at a gated eBook as our conversion point, we would calculate the conversion rate as the number of people who downloaded the eBook divided by the total number of unique visitors to the eBook’s landing page multiplied by 100.
If we had 100 unique visitors to our eBook landing page but only 40 completed the download form, then our conversion rate would be 40%.
It is important to note that we are using unique visitors and not overall visitors, as you do not want to double-count people who visit the page multiple times. Some people may visit the landing page twice before signing up, and you don’t want that to bias your measurement.
If you want your numbers to be accurate, however, you will also want to make sure you de-dupe sign-ups, as well. The numerator typically has much less double-counting and consequently should not cause as much of a concern for accuracy.
How to Calculate Conversion Volume
What is Conversion Volume?
Conversion volume is a measure of the traffic that passes through a given conversion point during a fixed amount of time. For this reason, it is commonly used as a measurement of the overall importance of a given conversion point in the customer acquisition process, as it measures absolute volume which can be compared relative to others or over time to measure marketing and/or sales performance.
Conversion volume is typically calculated as conversion rate times the total potential transactions at a given conversion point during a specified timeframe. However, it can also be calculated as the total number of completed transactions divided by the total time elapsed during these transactions. The formula you will want to use will depend on how you collect the conversion measurements.
Conversion Volume = Conversion Rate * Total Potential Transactions (Time is Constant)
Conversion Volume = Completed Events / Time Horizon
Example Business Application of Conversion Volume
For example, if we were looking at a prospect newsletter sign-up as our conversion point, we would calculate the conversion volume as the number of people who signed-up for the newsletter divided by the relevant time horizon.
In order to gauge the performance of this marketing channel over time, we would measure incremental conversion volume or change in conversion volume. Incremental conversion volume would be calculated as new conversion volume minus old conversion volume. Change in conversion volume would be calculated as incremental conversion volume divided by old conversion volume.
For example, if you had 500 newsletter subscribers at the start of a marketing campaign and 650 at the end of a two week period, but lost five subscribers during that period, what would be the conversion volume?
We can calculate total conversion volume as shown below:
The conversion volume would be 150 and the percentage change in conversion volume over the two weeks would be 30%. It is important to note we are using net new subscribers because subscribers will drop-off over time and we want to capture the net change. You don’t want this to bias your measurement upwards.
How to Calculate Cost per Conversion
What is Cost Per Conversion?
Very simply, cost per conversion is the amount of effort and money required to generate a conversion via a specific conversion point. One way to think of it is your return on investment, a measurement that can be used for budget analysis as well as comparing the effectiveness of different marketing and sales channels and tactics.
Cost per conversion is calculated as overall conversion cost divided by conversion volume over a specified period of time, where cost is measured in terms of dollars.
Cost per Conversion = Total Cost / Conversion Volume
In order to make this calculation, you will have to estimate the monetary cost of effort. The easiest way to do this is to assign a fixed cost per hour of human work. Make sure that this cost is consistent with the pay rate per hour of the team who is working to achieve these conversions. This will ensure that your estimates are at least within reason.
Example Business Application of Cost per Conversion
For example, you might want to estimate the cost per conversion in order to select the best events for your company to participate in next year, and determine how it should participate in those events in order to maximize your return on investment on event marketing. One way to look at this question is via cost per conversion.
Below is an example to demonstrate how you would utilize cost per conversion comparisons to make this type of decision.
Let’s say you are an event marketing manager at a B2B business and have been allocated enough resources to get involved in four events next year. How do you decide which events to get involved in and what type of involvement to have at these events to maximize your return on investment?
To answer this question, you probably want to evaluate at least 7-10 events in total and consider the different ways you could participate in those events.
Let’s consider that you participated in four events last year (two speaking engagements and two event sponsorships). To answer this question, you first want to identify/estimate some key information about the events your company participated in last year:
- Attendance of the events we participated in last year
- Event involvement level
- Time cost of involvement
- Number of attendees who signed-up for additional information
Using this information you will be able to estimate cost per lead created, as shown below:
By looking at the stats for last year’s event participation, you can identify the characteristics of successful events so that you can pick high quality new ones to consider for next year. Attendance, demographics, and competitor presence are three of the key attributes you want to evaluate. In doing so, you can identify the following three additional events to evaluate:
In order to compare these events against last year’s events, you need to identify event participation costs as well as attendance and estimate sign-ups and time cost of participation. From there, we can calculate an expected cost per conversion.
Once you have this information, then you can stack rank the events by cost per conversion to determine which ones you should participate in this year to target your specific buyer group.
This same decision could also be evaluated via conversion volume, but the big advantage of using cost per conversion is that it takes into account both financial and human capital inputs in the evaluation, where as conversion volume only looks at aggregate conversion volumes. Things like event participation can vary widely, so that would not be a realistic comparison of the events. In other instances, it could make sense to use this metric if the difference in financial and human capital costs is negligible.
How to Calculate Conversion Time
Now that you know how to calculate and apply conversion rate and conversion volume, let’s dive into how to calculate conversion time.
What is Conversion Time?
Conversion time is the average amount of time it takes per a conversion. It is commonly used as a measurement of how efficiently marketing and/or sales have been at getting prospects to move through a specific stage in the buying process.
Conversion time is the product of conversion rate multiplied by conversion volume or sum of total transaction time divided by the total number of successful conversions during a specified time period. The approach you take will depend on how you have decided to measure the conversion.
Conversion Time = (100 * Conversion Rate)/ Total Transactions
Conversion Time = Total Transaction Time/Successful Conversions
Example Business Application of Conversion Time
For example, if we were analyzing the turnaround rate for an email-based contract renewal outreach campaign, we would want to calculate the conversion time as the total transaction time divided by the number of successful conversions.
Hypothetically speaking, let’s say you had 50 customers with expiring contracts that you reached out to via an email-based contract renewal outreach campaign and seven of them agreed to renew their contracts. They took a total of 700 hours to do so. Another 13 rejected the offer and took an estimated 242 hours overall to get back to you.
Using this information, we can calculate average conversion time, as is shown below:
How to Determine Conversion Timeframe
The final piece of the puzzle is learning how to determine a conversion timeframe, evaluate your timeframe, and apply the timeframe to your business.
Determining Conversion Timeframe
Determining the appropriate timeframe to use is a critical step in calculating an accurate conversion measurement. This is because most conversion points have different time characteristics or seasonality that can lead to biases and/or meaningless calculations. Consequently, it is a factor that you must pay special attention to when calculating conversion measurements. The only time when time will be an irrelevant consideration will be when you calculating conversion rates or cost per a conversion for an instantaneous conversion like a pay-per click advertisement.
Evaluating Conversion Timeframe
Think of conversion analysis as an experiment where you are trying to control outside factors from influencing your measurements in either period either by balancing the impact in each time period or by removing the impact all together. The time horizon can be used to do this or allow for you to remove the impact via a correction factor.
For example, to remove a seasonality effect, you will either want to confine the effect to a single period or have the same exposure to the affect over both periods you are measuring. If you are trying to confine the effect, it is best to have the time period either inside of or outside of a seasonality period, not overlapping it. This will allow you to develop a more precise correction factor.
Some rules to consider when evaluating timeframe:
- Must be longer than the maximum expected conversion time to get a reliable estimate.
- Should not be longer than the expected impact of the marketing or sales tactic or strategy you are looking to evaluate.
Example Business Application
A marketing manager at a B2B software firm is trying to identify the conversion time for a critical software upgrade that it pushed out to its customers. The marketing manager believes that the upgrade conversion time is an indicator of how important the software is in their overall business. He wants to use this information for helping him identify customer segments that will help them increase average customer tenure and the average customer lifetime value.
He knows from prior releases that his customer base had taken an average of 2.5 weeks to upgrade the software with three push notifications, but knows that the product is used daily by his best customers. This has him concerned about customer retention problems down the road. The product is too new to measure this through actual churn.
He knows that users who take more than a month to upgrade are outliers, so he believes the maximum upgrade timeframe would be a month. He knows this excludes a few very late upgrades, but determines this is better than introducing the biases related to having too long of a timeframe and also introducing the bias of including part of holiday season in calculation. Consequently, he decides to calculate the number of successful user conversions during that month and track the conversion time for each user. That will allow him to then calculate conversion time using the following formula: Conversion Time = Total Transaction Time/Successful Conversions.
After one month he measures 800 successful conversions and a total transaction time of 232,128 hours. He also noticed that 50 users never upgraded.
Using this information, he is able to calculate the average conversion time, as is shown below:
When he takes the number back to the management team they tell him that is great but this number is meaningless — they need a measure on a per firm basis for segmentation not a per a user level and need to categorize the results by industry. To correct this issue he then does the calculation per a firm using the same formula, but has to average the statistics for each firm.
To make these calculations, he assembles the following table below:
Using this industry level average conversion time measurement he determines that marketing agencies are the best types of firms to be targeting and takes this recommendation back to his team.
Conversion analytics can be used as a decision support mechanism throughout your business. However, recognizing the potential applications and how to leverage conversion analysis to derive decision making information is sometimes not as straight forward. My hope is that these examples will get you started looking at the different ways conversion analytics can be applied in your business and teach you how to go about thinking about these calculations, so that you can better leverage the data available to your organization to make decisions with less uncertainty.
Now you know the ins and outs of calculating conversion measurements you can start using these calculations to make all kinds of decisions!