Real Time Experiment – Testing the Muse in Practice (Part 2)

This is Part 2 in the series of Testing the Muse in Practice, our first real time experiment. Recall from Part 1, detailed in this blog post, that we started with a personal problem that needed solving:

Comparing multiple used cars apple-to-apples is impossible.

To assist in the process, I researched what others have done to combat this problem, and found a potential solution:

A forum comment that described a simple comparison calculation for ranking similar used car prospects.

Tweaked, customized and entered into a spreadsheet for my own use, I now had a systematic basis for comparing similar vehicles, to get an idea of which dealers/sellers provided good value or which were pricing their vehicle too high.

It served its purpose, and helped me select a family vehicle from a list of 4 prospective vehicles. Then I promptly filed it away, hoping never to have to buy a used car again for a very long time!

But just this August, during an intense period of inspiration where I was going through older notes I had made, it dawned on me that I had this spreadsheet in my possession – something that could quite potentially be of serious value to other people; so I dusted off the document and started brainstorming…

In this post, we’ll begin to examine market research hacks – attempt to identify the active seeker for this potential product, and determine if enough of an audience exists, in order to proceed to Testing the Muse.

Do Others Have This Issue?

As I mentioned in the last post, if you have a “problem” (no matter how big or small), it’s very likely others have this exact problem as well. And where there’s a problem, there’s someone willing to pay for a solution.

Before you can bank on someone’s willingness to pay for a solution, you first need to find out where that “someone” is – who are they, where can you find them, and how many of them are there?

With this particular scenario, we’re looking to identify people who are:

  • Searching for used cars; and
  • Having trouble comparing options and making a decision.

Wow – that’s broad.

Market Research

It’s important to have an understanding of your potential audience – as this will influence certain advertising efforts when later Testing the Muse.

So, at first glance this audience appears to be pretty much anyone who is old enough to drive. Gulp! However, for our purposes, we could reasonably refine the demographic down to target:

  • Both males and females (since both drive and buy vehicles);
  • Located in the United States (largest used car market, a good starting point);
  • Aged 18 to 65. We’re eliminating anyone younger, as they either can’t drive or wouldn’t likely be purchasing on their own; and excluding anyone older than 65 because, well, we have to draw the line somewhere;
  • Average income (financial ability to afford a car);
  • Actively searching for used car ads online (duh!); and
  • Yada, yada, yada…

I’m going to stop right there.

You could spend hours, days or weeks doing Census research, building a demographic and socio-economic profile, and cross-tabulating data variables of age, income, vehicle ownership, etc. I know the amount of work (and expense) involved in this type of data gathering and analysis, because I’ve done it extensively for well over a decade in my dayjob.

For certain commercial purposes, there’s value in that – if you were preparing a Business Plan. We are not.

We’re going to skip over trying to get a big picture of the market – because likely only a small fraction of the total number of people searching and purchasing used cars will want our product. We’re going to hack the market research and go straight to finding out who is our active seeker – who’s interested in this topic and who are actively searching for the answers we hope to provide.

We can complete this hack with two widely available (and Free!) sources – taking advantage of the two of the top tech savvy companies in the world, that also happen to collect the largest number of data points about consumer behavior: Google and Facebook.

Quickstart Market Research: Finding Potential Customers

Although I still refer to it as “market research”, we’re actually scoping our research down to specifically ‘finding potential customers’. It’s more realistic, provides better insight, and is actually easier to perform.

Brainstorm Keywords

Keyword research – some people love it, some people loathe it. The fact is, if you’re doing work online, you need to know the basics. I actually didn’t enjoy it that much to begin with, but now with tools like Long Tail Pro, it’s become both easier and more efficient.

This is far from a tutorial on keyword research – it’s just too big of a subject to do justice in this post, but it will give you a brief view into [how I perform] the initial stages of this process.

The first step is to brainstorm potential words or phrases your ideal customer would be searching on the internet. For example, people may be searching for: used car research, tips for buying a used car, comparing used cars, etc.

Like all of my brainstorming, I just write or type everything and anything that comes to my mind. In this instance, I just have to mentally put myself back into 2012-13, when I was actually in the position of trying to find a tool to help me compare used cars. What words did I search for? If I were to do it again, what phrases make sense, and what phrases would lead me to a product that I’m currently investigating?

The task here is to get as specific as possible, because a broader search term (e.g. “used cars” or “buying a used car”) are too broad, will be too competitive, and will ultimately give you poor results.

You don’t have to go too crazy here, because we’re going to validate and refine our keywords using actual data in the next step. But I firmly believe you should mentally and physically do the brainstorming first, before diving into the keyword research. It helps put you in the right mind-frame, so that you can “think” about yourself as a customer first.

Active Keyword Searches

Once you’re ready to start looking at the numbers, take your list of potential keywords and start plugging these into Long Tail Pro (paid) or Google Adwords: Keyword Planner (free) to observe the number of active searches for these terms.

Google’s Keyword Planner is a free tool within Google AdWords that helps you, among other things, research keywords and ad group ideas for statistics on search histories and performance (i.e. how many people are using those keywords while searching online). Long Tail Pro, on the other hand, is paid keyword research software that allows the user to generate unique “long tail” keywords (a phrase of 2+ words) based on a your primary targeted keyword(s). Long Tail Pro could be used as a feature rich replacement to the Google Keyword Planner, or as a powerful supplement to your Google keyword research.

These tools will also provide a huge number of related keywords, which is why I suggested not spending too much time on brainstorming phrases above.

If you prefer, you also have the option to start by defining where you want to focus your efforts. For this exercise, I’d like to specifically see data on searches in the United States, in English, excluding words (“Negative keywords”) such as “sale, selling, price, cheap”, as well as plural versions of these words.


I chose to start with one specific keyword phrase: comparing used cars. Well, it turns out “comparing used cars” is Terrible!

It only has 30 average monthly searches (you want this preferably in the thousands!) and the competition for this keyword is “High” (you’d prefer “Low”). We’re going to have to dig deeper here. Sort the keywords (by column header) and attempt to find phrases that have a high monthly search average, with low competition – it’s not hard work, but it can take a bit of time – in total, the Keyword Planner identified 563 related keywords!

To illustrate how powerful keyword tools can be (and how what you “think” others want, isn’t necessarily always true) – I merely tweaked the word “comparing” to “compare” and instantly the results looked considerably more favorable: 1,300 avg. monthly searches, with Medium competition. Still not spectacular numbers, but definitely heading in the right direction.

On the flip side, if you were to generalize your keywords too much, such as using a broad term like “used cars”, you’d get results like 450,000 avg. monthly searches, naturally with significantly higher competition.

As you can see, generalized terms yield huge numbers. Too big, in fact. This is why you should always aim to narrow down your keywords as much as possible. Otherwise, when Testing the Muse using online advertising later on, our costs will be astronomical.

There are a few techniques to making your keyword research more efficient:

  • Utilize the “Negative Keywords” function – i.e. enter “sales, sale, selling, sell, prices, price, cheap” as Negative Keywords, because we are not selling or searching for used cars to buy (the majority of searches) – we are trying to find a tool to help us decide on which used cars to select. Remember the 563 keyword suggestions? That would’ve easily been doubled if we had not excluded these words.
  • Favor phrases (2+ words) over single words. They generate more specific search results and are generally less competitive. Long Tail Pro is highly recommended for making this extremely easy.

The domino effect is evident: the more you can niche down your market, the better you can identify your customer, the better you can focus your advertising (and offering), the less money you waste, the better the results you will get, and so on.

Although the keyword research is yielding pretty poor results, I’m not that concerned for a few reasons:

  1. We haven’t spent a penny yet – no harm, no foul;
  2. This is simply research, it’s not a predictor of future results – for that, we must test. Theory is good, data is better, but action is proof; and
  3. We’re going through this process to investigate ‘potential’, or at the very least, for practice; our livelihoods do not depend on the success of this yet-to-be-determined product.

Nonetheless, we still want to test to see what’s possible, with the aim of success – so let’s move ahead…

In the next post, we’ll continue our “market research” with two additional tools designed to help hack the market research process, and bring you closer to identifying and finding your potential customers.

Don’t miss the continuation of this over-the-shoulder view of a potential product-in-the-making – sign up below!

Key resources mentioned:

My next to-do:

  • Continue to refine the market and build on the customer profile developed for this concept.
  • Prepare Part 3 of this real time experiment.

Action you can take:

  • Be the first to get updates on this process and see Part 3 of this real-time Testing the Muse experiment by subscribing below.
  • Show this blog some Love: If you have contacts that you think would enjoy this real time experiment, please share this post!



2 thoughts on “Real Time Experiment – Testing the Muse in Practice (Part 2)

  1. […] an estimated audience of approximately 1,750,000 (identified during the market research stage – Part 2 and Part 3) and now we have a dollar figure (literally $1) to assign to the product. Since we now […]

  2. […] already done diligent work when researching your target keywords (Part 2 of Testing the Muse in Practice) using Long Tail Pro and Google AdWords Keyword Planner – this helped you determine your […]

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes:

<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>