Understanding and Measuring the Impact of Podcast Visibility Optimization (advanced version)

We recently published our easy guide to measuring the impact of podcast visibility optimization. With this guide, we wanted to empower everyone to understand the relationship between PVO and download data—even those without a data analytics background.

Today, we want to take things further. This article is for people who wish to deepen their understanding and take their podcast data analytics skills to the next level.

It will be especially useful if you run several marketing channels parallel to PVO.

Here are the topics we will cover in this article:

Table of contents

Note: This article assumes you are already familiar with concepts like PVO and have read the easy measurement guide. If you need help understanding any of the terms mentioned in the piece, please refer to our podcast glossary.

The importance of Measuring PVO

It can be tempting to just apply best practices to your metadata once, cross it off your list and call it a day. In reality, PVO is a continuous process.

As such, it requires regular monitoring and measurement. How is your position evolving for a given keyword? Are new competitors encroaching on your turf? Are there any changes to the algorithms that require adjustments? The list goes on and on.

Opportunities to rank on new long-tail keywords will arise as you publish new episodes. Regularly using your most relevant keywords in your episode titles will also increase the relevancy of your podcast for these terms at the show level. Doing so might give you the extra boost you need to capture a top position for a high-volume search.

You need to measure the efficacy of these sustained efforts.

It also serves as a motivational tool for your team. If your producers and editorial team are going the extra mile to write search engine-friendly episode descriptions, seeing the positive impact of their work can only further motivate them to do a good job.

Last but not least, it helps in reporting accurate results to management. If you spend time and effort on PVO, you will want to put a number on it and prove that these resources are well invested.

What tools can you use to measure?

If you are familiar with SEO, you know that Google Search Console provides a wealth of data. It can tell you what search requests brought people to your site and how many visitors your SEO efforts generated.

Unfortunately, there is no such thing in the world of PVO. Apple and Spotify don’t offer direct measurement options.

However, combining an advanced podcast analytics tool like Voxalyze and Excel will give you a fairly accurate estimation of PVO-generated downloads. We will tell you how later in the article.

PVO measurement tips and best practices

Accurate measurement starts with solid data foundations.

Create a solid data infrastructure

Your data infrastructure needs to be ready for measurement. Your hosting provider’s built-in analytics will only bring you so far. To accurately measure trackable marketing channels, you must use a more advanced audience analytics solution, such as Voxalyze or Chartable.

Know your data in and out

Before optimizing your metadata, we recommend you get a clear picture of your downloads.

Looking at your historical data, you need to answer the following questions:

What has my month-over-month (MoM) download growth been in the last 6-12 months

Knowing this will help you identify a baseline trend (if any). For instance, let’s assume your downloads have increased by about 10% every month in the last quarter. Now you start optimizing your metadata, and your MoM growth jumps to 20%. In that instance, PVO only accounts for a 10% increase (20%-10%).

Of course, this simplified example assumes all the other parameters remain unchanged. In reality, you need to understand which levers were fueling your growth before PVO and monitor any change to them. For example, if you ran paid campaign for months and stopped just before starting PVO, it would be reasonable to attribute the full 20% to PVO.

What was my month-over-month download growth last year?

Knowing this will help you understand and factor in the role of seasonality. Let’s assume you start optimizing your metadata in June. As we will see in the article, you can expect to see the first results in July. But July is also the beginning of the summer slump. You usually observe a 30% decrease in downloads MoM. This year, your downloads decreased by 10%. Should you conclude that PVO doesn’t work? Of course not! It actually helped dampen the shock.

Can I expect more variability due to other factors?

For instance, do you have a famous guest lined up?

If there were outliers in my data, how can they be explained?

Was there a measurement bug caused by my hosting platform? Did an event in the news cycle or a Netflix release cause a sudden rise in popularity for my podcast’s topic?

Will I maintain my publishing pace?

When comparing periods, a difference in the number of new episodes released can significantly influence your download numbers. Overlooking this factor might lead you to under- or overestimate the impact of PVO on your downloads. To compare things apples to apples, you need to calculate prorated numbers.

Let’s look at an example:

  • Reference period: 5 episodes released, 10000 downloads
  • PVO test period: 4 episodes released, 9000 downloads

Had we only considered the total number of downloads, it would look like a 10% decrease. In reality, our downloads per episode increased by 12.5%.

Prorated downloads per episode

You now have the keys to think about your data critically. With all these variables in mind, you can build a model with your BI team and estimate the number of downloads you can expect in the coming months. In the rest of this article, we will call them your projected downloads.

What are the primary and supporting metrics you should track?

Primary KPIs

  • Monthly downloads and MoM growth
  • Average number of weekly downloads
  • Minimum amount of daily downloads. Take the average of the 4-5 weakest days in any given month. Doing so will prevent outliers from skewing your data. By outliers, we mean days with exceptionally low (or high) download numbers that deviate too much from your usual highs and lows.
  • Maximum amount of daily downloads. Same as above, but for the 4-5 days with the highest number of downloads.

Supporting metrics

Visibility score and rank

Voxalyze’s visibility score (VS) helps you understand how visible your podcast is when you compare it to all other shows in a given country and platform.

Keyword positions

Monitor your ranking for your most important keywords. We recommend you check it bi-weekly and after making significant metadata changes.

Keyword count

This is also an interesting metric to watch. Its main role is to assess the quality of your episode metadata. If your editorial team applies PVO best practices correctly, you should be able to rank for an increasing number of longtail keywords.

The Daily - Supporting visibility metrics
The Daily - Supporting visibility metrics

By the way, did you know you can create a free account to see the visibility metrics of your own podcasts?

How much time before seeing results with PVO?

The timeline to witness the first results of PVO can vary based on different factors like your podcast authority, frequency of episode releases, or your promotional strategies.

With all these caveats in mind, and assuming the team is quick to embrace PVO best practices, it’s common to start seeing initial results within 2 to 3 weeks.

PVO usually works in 2 phases.

Phase 1 corresponds to the first round of metadata updates. Adjusting the show title and description provides a first, pronounced visibility boost that translates into increased downloads.

Phase 2 starts after that and lasts for as long as you keep optimizing your metadata, episode after episode. The visibility curve usually looks like this:

Typical PVO growth curve

How to calculate how many downloads are generated by PVO vs. other channels?

Assuming you did your keyword research well and are optimizing your metadata consistently, the download uplift generated by PVO is generally substantial enough that it becomes evident that your PVO efforts are paying off.

Our audience development case studies are a testament to that.

APM Studios - 24% audience growth in 4 weeks
APM Studios - 24% audience growth in 4 weeks

The trickier question is: how do we distinguish PVO’s impact from other marketing channels?

When you see that download graph going up, how many incremental downloads can you precisely attribute to PVO?

After all, you’re also promoting your podcast in your newsletter, on social media, and you run cross-promotional campaigns in other podcasts in your network.

Well, first, were you running these channels prior to launching PVO? If that is the case, downloads coming from marketing campaigns are already factored into your projected downloads.

The formula you want to use is straightforward:
PVO downloads = actual downloads – projected downloads

What if you start running new marketing channels and PVO in parallel?

To answer that question, we need to distinguish between two primary use cases:

  • trackable channels
  • non-trackable channels

Trackable channels

First, let’s define what a trackable channel is. Trackable channels are promotional methods that require users to click on a link or scan a QR code to access your content. Typically, your newsletter, website, and social media posts (with notable exceptions like IG posts).

Using a tool like Voxalyze’s podcast marketing analytics, you can use a tracked link to match clicks with listening events. Doing so will give you an accurate picture of the number of downloads from tracked channels. 

If you’re only running these types of channels on top of PVO, you can calculate the impact by using this simple formula:

PVO downloads = actual downloads – projected downloads – downloads from tracked channels.

Non-trackable channels

This type of channel traditionally includes TV, radio, PR, and in-person events. If you advertise your podcast on TV, the only action a viewer can take is to visit your website, download your app or type the name of your show in the search bar of their favorite podcast platform.

There is no banner to click. No QR code to scan. No tracking link loading in the background.

As a result, tools like Voxalyze can’t map ad interactions with downloads. Measurement becomes more complex. We cannot use the simple formula mentioned above. Let’s see several methods to solve this.

Separate testing periods

The easiest solution consists in refraining from promoting your podcast on non-trackable channels during the first six weeks after the first metadata changes. Doing so will reduce the noise in the data, help to tell apart downloads generated by PVO, and establish a new baseline. We understand this is not always possible, so we provided more measurement methods below.

Force tracking

You can try to mitigate the absence of clickable elements by incentivizing users to use a vanity URL to unlock exclusive content. When the page loads, tracking is triggered. However, this method is insufficient in and of itself. People are lazy, and many will simply google the show’s name. Vanity URLS will only slightly increase the amount of tracked users.

The peaks and valleys method

We explained this method in our beginner’s guide. It is fairly easy to implement and estimates the PVO uplift by comparing days in-between episode releases when downloads are at their lowest.

Geo-based lift measurement

When running a test campaign on an untrackable channel, a best practice consists in allocating your budget only to specific areas of a country. For instance, you could promote your podcast on local radio stations in California.

The purpose is to split your audience into a test group (Californians) and a control group (people in the rest of the US).

At the end of the campaign, you compare the outcomes between groups to evaluate the effectiveness of the channel. Did the increase in downloads from California justify the budget you put behind the campaign?

The great thing about this method is that we can piggyback on it to measure the impact of PVO indirectly. How? Let’s go back to our example campaign in California.

Let’s say you launched that campaign around the time you started optimizing your metadata. You see the following numbers:

  • California: 30% increase in download
  • Rest of the US: 15% increase in download

Since you are sure that people outside of California didn’t hear your radio advertising, you can ascribe the 15% gap to PVO.

Even better, you can now conclude that your radio campaign contributed to a 15% (30-15) increase in downloads, as opposed to 30%. This sheds an entirely new light on the ROI of that campaign! Is your cost per download still within an acceptable range?

This is a simple example where we assume all other variables remained equal. In reality, other parameters will always deviate a bit, but you can take steps to reduce their influence to a minimum:

  • Chat with the editorial team beforehand to make sure they can maintain a stable release calendar.
  • Don’t launch such campaigns around periods with high, seasonality-driven fluctuations
  • Consult your BI team to ensure the size of the test group and budget create the conditions to reach statistical significance.

Want to measure the impact of PVO and other channels on your downloads? Sign up and create your free Voxalyze account now.