Your Conversion Rate is Wrong

There is a large issue with noise when it comes to Google Analytics in the real world. “Noise” is what I call any traffic or conversion that could not possible be a customer in your business model.

Testing and analyzing traffic uses the conversion rate as the golden metric, but what if you convert a lot of butchers on the West Coast when you are a restaurant in Brooklyn? Your website will show that you got a lot of conversions and that you should do more of the same advertising, while your staff are tired of the butchers calling in to the shop, trying to sell you meat.

An example from my own work is when I consulted for a web development company doing contract work. Their website got a lot of inbound traffic, all of which generated “conversions” through their contact form. However, a lot of those “conversions” were from ‘vendors’ in India, looking for business from them, not customers (deemed the “India” problem).

This was further compounded by Indian vendors finding them online and calling in to the office, tying up time that could be better spent on business development efforts. This left them politely declining new business “opportunities” and less time to pursue real revenue. The noise they created online turned in to noise in the office and lost salaries.

This same thing applies to your site — traffic is great, but what you really want are more customers. This means you need to know where your customers are coming from and gain more sources of that traffic. This also means ignoring traffic that doesn’t contain customers (even if that traffic contains a lot of conversions).

We do this by segmenting and filtering. By removing noise, the people that are not going to give you money no longer count in your conversions, which gives you a conversion rate you can actually use!

Clarify Your Customer

There is a lot that we can do to segment out customers, but to start with, let’s focus on the following:

  • What language do your customers speak?
  • What is their location? Country? Region? City?

Even if you only account for these two things, you will remove a large swathe of your noise.

Segmenting Traffic

Creating new traffic segments is a great way to compare what your ideal customers are doing to what the rest of your visitors are doing. This will help you evaluate if your ad spend and marketing is worth it, which sources of traffic are better than others, if you have specific pages on your site that are bringing you tons of conversions but no customers (and requiring your business to spend time personally declining).

Creating segments also allows you to try new things without effecting your analytics in any serious or permanent way, as well as account for multiple customer profile types. How does an older demographic compare to a younger? What attracts each type? That sort of thing.

Creating a New Google Analytics Segment

First, go to Google Analytics and create a new segment. It starts out as all of your traffic, so you can whittle down from there. Based on what we considered before, you will want to specifically check out people that come to your site speaking your language and that are in your locale.

Selecting Language and Location in Google Analytics Segment

Remember, you are trying to segment out people that don’t fit your generic customer profile. I am aware that someone in New York could visit your hair salon in LA, but you are much better served focusing on growing your traffic in LA and I think you would agree.  So, look at where you are specifically getting traffic from people in LA and do more of that.

Filtering Traffic

Creating and applying filters is a more permanent approach. When you create a filter, traffic that doesn’t fit never even makes it to your analytics board. This is especially useful when you know that you don’t want specific traffic to be included in your view.

Make sure that you do this in a new ‘View’ in case you aren’t happy with the results. Once you are happy with your set of filters, you can re-apply them to your master view.

Checking Test View and Selecting a Filter

The classic example is applying a filter to remove traffic from your own computer, which is what the “Predefined” filters are for. However, it can be helpful to take it further and remove traffic from places around the world that you don’t operate. 

Predefined Filter in Google Analytics

To instead filter out traffic from particular regions, you will want a “Custom Filter”, which changes the entire look. Once you select “Custom Filter”, you will be able to choose from a (huge!) list of possible filters to include or exclude.

Custom Filter in Google Analytics

Reducing the “noise” in your analytics will allow you to spend your time and money on efforts that are actually gaining you new business and new revenue while allowing you to quickly diagnose and limit any problems that suddenly bring you a lot of new conversions, but no new customers.


Looking for additional help? Questions? Leave a comment, send me an email or find me on Twitter. If you need more personalized help, let’s set up an initial consult to find out where your problems lay and what we can do to make your online presence even better.