What I learned about Google Adwords by spending $103.9

During February, I have set myself a goal of understanding how Google Adwords work by testing them for a few products I sell. I have allocated a budget of $100 for this purpose and set out to create my first ever campaign in Google Adwords. This article tells you the story of how my experiment turned out.

Specifics about the Ads I ran:

I ran ads for 2 of my products – Excel School & Project Management Templates. Both of these sell quite well. The sales pages convert about 3% of visitors (about 1.2% if you calculate conversion ratio based on page views).

  • The ads ran between February 9th and 27th.
  • The expenses were – $61.2 for Excel School & $42.7 for PM Template ads

Experience with Google Adwords:

Ad Creation & Setup:

It took me a while to figure out Google’s Adwords website. It looked cluttered and confusing to say the least. May be I am not qualified enough to use the site. After spending a couple of hours, I could figure-out quickest way to setup ads, adjust budgets and start the campaign. The details of particular ads I have ran are below:

Ad creatives:

  • I have designed 4 ads in total, 2 for Excel School and 2 for PM Templates. The ads are shown below:
  • I have specified that ads should be shown uniformly over time (ie, each ad get same number of impressions).

Targeting & Budget of the ads:

  • Keywords: I have used a dozen highly targeted keywords that are relevant for each product. I am omitting specifics as they are not relevant here.
  • Geographical Targeting: I have analyzed my sales and based on that data, I have targeted most of Western countries.
  • Search vs. Display Network: This is Google’s way of asking whether you want your ad to be shown only in search results or even on sites that have adsense blocks. I initially set this to everything, later (I guess after 4 days) changed it to Search only.
  • Device Targeting: I have specified that ads should be shown only computers (ie not on mobile phones etc.)
  • Budget: I set the budget to $10 per day for Excel School & $4 per day for PM Templates
  • Bidding: This is Google’s way of asking how much would you pay for each click. I set the maximum CPC (Cost per Click) to $1.00 for Excel School and $0.75 for PM Templates.

Sidebar: How much budget is appropriate?

Since I was running a one-off experiment with the ads, I just set my total budget to $100. But in general, it is prudent to decide budget based on your daily sales & conversion ratios.

For example, if your sales page conversion ratio is 3% (ie for every 100 visitors, you sell 3 units), and your profit (sale price – cost) per sale is $25, then, for every 100 visitors you make $75.
That means, you cannot spend more than $0.75 per visitor to acquire traffic.
So this should be your maximum budget.

In fact, your budget should be much lower than that as you will have other expenses (for eg. taxes, shipping, customer service, returns, website hosting etc.)

How did the ads perform?

My Adwords results were mixed.

  • Google Adwords sent me 63 clicks for Excel School & 64 for PM Templates. (Note: These numbers are too small to conclude anything statistically)

Since I use a 3rd party shopping cart system, measuring exact conversion ratios was not possible (at least, I did not know how to). So I just used a crude approach.

ie. How many sales did we make?

  • Between 9th and 27th Feb, I sold 76 units of Excel School & 95 units of PM Templates.
  • There were 4650 and 7880 page views to these products respectively.
  • So the conversion ratio was – 1% and 1.2%.

Since I had no idea how many of these 63 and 64 clicks converted to sales, I just assumed that “Ads convert 3X better.

Mind you, this is a highly stupid assumption that favors Google. But I went with it anyway.

So, with that logic, I sold 63*1%*3 = 1.89 Excel School units and 2.3 units of PM Templates.

If I take the sale price of these 2 products ($97 average sale for Excel School and $30 for PMT), we get,
$187 and $72 as revenues.
Now, my Adwords expenses were, $61 and $42 for both the products
That means, I had a notional profit of $126 and $30 on these products.
Mind you, I have not included the actual product cost in this calculation. If I did that, my profits would be even less.

Detailed report of Adwords Performance:

Google Adwords Performance Report
Google Adwords Performance Report

What is the real truth?

I was not satisfied with the above calculations. So I explored my Google Analytics to understand how the visitors from Google Adwords behaved.
Wait for the surprise…

Google Adwords Report from Google Analytics
Google Adwords Report from Google Analytics

(Note: For some strange reason, Google did not track the visitors it sent to my site in the first 4 days. I later enabled the setting in my Adwords account. Only then, I could find the adword visitor details in Google Analytics reports. I suggest you double check these settings to avoid shooting in the dark.)

As per Google Analytics, I had 20 visits to Excel School & 35 visits to PM Templates pages.

Now comes the kicker. Out of the 35 visits PM template received, on an average each person spent 4 seconds on that page.
There is no way anyone could have made a purchase decision in such short time. In other words, I am sure, I did not sell even a single unit of PM Templates to adwords visitors.

The results for Excel School were encouraging. On average, each visit laster 2 and half minutes. I am sure there are 3-4 visitors who spent much more than that. And may be, 1 of them would have joined Excel School.

Bottom line: Conclusions on Google Adwords:

  • Unmatched Exposure to your products: Even though, there are only 120 odd clicks, As per Google’s report, my ads got a whopping 290,000 impressions. Even if only 1% of these impressions were noticed by actual prospects, that is 2,900 people. I am impressed.Google Adwords - Ad Statistics
  • Costs not in favor of small businesses: Again, this is not based on any statistical proofs. Based on the amount I spend, time involved in setting up Google Adwords and managing it, I find that Adwords is not that promising. It might work well, if I can figure out the magical combination of right keywords, targeting, CPC bids and sales page layout. But then, often, start-ups and small businesses are too busy doing other important things, like marketing their products, working on product features etc.
  • Do not advertise if you cannot measure: That is right. If you cannot measure your conversion ratios, there is no point advertising. I am in the process of setting up an online store so that I could track conversion ratios, run split tests and simplify purchasing process for customers. Once it is ready, I can re-run the adwords experiments with more confidence.
  • Mysterious Keyword Quality Score from Adwords
    Mysterious Keyword Quality Score from Adwords
  • Do not target 6Bn people: simple, you will never sell by targeting entire population. Instead, focus on a handful of specific keywords that convert very well. As an aside, Google has some strange mechanism called as Quality Score for your keywords. Again, unlocking the mystery behind this is an art and Google Adwords website is like a maze. You could spend 2-3 hours without doing much there.

If Google Adwords is like this, then what is the best way to sell?

Of course, there are cheaper and more effective ways to sell online. Start with these,

More: 7 ways to sell more while keeping expenses low.

What is your Experience with Adwords?

Do you run adwords campaigns frequently? What is your experience with it? Do you find the returns on investment good enough? What tips and ideas can you share with our readers?

Please comment and share with us

PS: I also ran similar experiments with Facebook Ads. More on this in the next post.

PPS: Also read Patrick’s take on Adwords. His blog is a recommended read if you are running web-based startup.