Do I need to advertise on my own company's name?

Many PPC marketers doubt about the rentability of PPC keywords who contain their own company name. Suppose you have a small furniture business named "JohnJohnsonElectro.com". Every search term who contains the word "JohnJohnsonElectro" is in the so called group "branded keywords". People who search for those search terms, know about the existence of your company and are already searching for a product inside your website. It depends on your brand or company name how big the search volume of branded keywords is. Since one for these search words is on the first page of the organic search results anyway, if your SEO is well organized, you can ask yourself if it is really necessary to place a paid ad. One way to do this well, is by drafting two campaigns; one for all branded keywords, so with your own company name, and one for all non-branded keywords, so without your own company name. This way you can easily measure and keep an eye on the rentability of the two campaigns in Adwords. Almost always has the branded campaign a much higher ROI than the non-branded campaign. Of course this is logical! However, the question is if these data do make up a good image of the rentability of those campaigns. Does Google maybe sketch a too beautiful image?

The "Last-Click" principle

Google Adwords calculates a conversion according to the "Last-Click" principle. May someone have come to your site by more advertisements before converting, the conversion is calculated on the last clicked advertisement. What does this mean for the total conversion data of the branded and non-branded campaign? I will give a further explanation by giving you an example.

Suppose: someone enters the searchterm “drill” in Google and clicks on your ad. Because he did not search for your company name, this search word will come from the non-branded campaign. The client browses through your website and leaves it. After a couple of minutes the customer realises JohnJohnsonElectro.nl has the best offer and enters the searchterm "drill JohnJohnsonElectro" in Google. Now the advertisement of the branded campaign pops up. The customer clicks on it and converts on the site.

Because of the "Last-Click" principle this conversion is included in the branded campaign by Adwords. However, this campaign might not have done most of the work. Probably the customer had returned to your site anyway, only the decisive click was the click on the first ad of the non-branded campaign! Anywise, Adwords now gives a too optimistic image of the ROI of the branded campaign and a too pessimistic image of the non-branded campaign.

A research on the effect of Last-Click conversion attribution.

Read more about this statistic.

Now the question is: can we quantify this problem? For example, is it possible to overtake the profit from this type of conversions and to transfer it from the branded- to the non-branded campaign. This is possible by combining data of Adwords and Analytics. We will use the Multi-Channel-Funnel Reports of Analytics and the profit data of Adwords.

How you do it

First, we will determine the costs and reached profit of all branded and non-branded keywords of last year. For this, go to your Adwords account and go to the keywords tab page without selecting a campaign. Select a timeframe, in my case, a year.

Set the time frame on a suitable period with enough data.

Thereafter we will enter a filter, who will filter all branded search words.

Enter this filter.

We note all total costs and profit data.

We note the total costs and profit data.

The total costs and reached profit of the two groups of search words is like this;

• Non-Branded Keywords:
  • Costs: €14.353,-
  • Profit: €100.182,-
• Branded Keywords:
  • Costs: 6.017,-
  • Profit: 259.878,-

From this follows a ROI of 6.98 for the Non-Branded Keywords and 43.19 for the Branded Keywords. We are now going to find out how big the part of the profit is which comes from the scenario shown above. So, a customer who first clicks on a non-branded ad, then undertakes other actions, clicks on a branded advertisement and then converts.

Open up your Analytics account and enter the just used timeframe, so again, a year.

Enter the same time frame again.

In the left menu, go to the tab page "Conversions" -> "Multi-Channel Funnels" -> "Top Conversion Paths". In this window we can, with some work, specify the foregoing scenario and determine the associated profit. First, the conversion type needs to be set. Make sure all conversions are selected who are also registered via Adwords, so for example the purchase of a drill.

Select the conversion types.

Now we will go and play with the so called "Channel Grouping". Because the preset is too rough for our problem. We prefer an subdivision in "Branded PPC Searches", "Non Branded PPC Searches", "Branded Organic Searches" and "Non-Branded Organic Searches". By creating these Channels and then specifying the right channel-order with a filter, we have the report we are looking for.

First click on "Channel Groupings" and next choose "Copy Basic Channel Grouping template".

Click "Copy Basic Channel Grouping template"

A new venster opens up with the title "Create or edit channel grouping". Herein, the eight standard groups are already entered.

A new venster opens up with eight presetted groups.

We can already delete the groups "Paid Search" and “Organic Search”. In my case the group "Display" is not necessary so I deleted that one in the example as well. Also name this template, for example "Branded vs Non Branded". You can do this on top.

Erase the groups "Paid Search", "Organic Search" and "Display".

Then click under on "+ Add new Rule".

First we will be creating the group "Branded Organic Searches". It is smart to number the titles of our self-made groups so they will be easier to filter. Give this group the title "1 Branded Organic Searches" and enter the first condition as in the image: "Include medium containing organic". Click on "Add 'AND' statement" and enter the following condition: "Include Keyword Containing [company’s/brand name]". (Fill in your own company or brand name for [Company’s/brand name] !). If your brand name can be written in more ways, or if there can be many typos you can add all these variants with "OR" statements. In my case I have added five different variants. Give a color to the group and click save.

Enter the first group.

We will enter the following group "Branded CPC Searches". Enter the title with number 2 in front. The design is the same as with the group "Branded Organic Searches", only now we will enter as a first condition "Include Medium Matchin RegExp ^(cpc|ppc)". See the image below. Then comes an "AND" statement followed by your company name. The company names are all separated by "OR" statements.

Enter the second group.

The next group will be "3 Non Branded Organic Searches". Enter as a first condition "Include Medium Containing organic". Just as before, it will be followed by the variants of your company name: "Exclude Keyword Containing [company’s/brand name]". Pay attention! Right now they will be separated by "AND" statements!

Enter the third group.

The last group will be "4 Non Branded CPC Searches", analogue to the "Non Branded Organic Searches". Enter the data as shown in the image below.

Enter the fourth group.

We will now get the following image.

The report with the new groups.

Of course the colors can deviate depending on the choice by entering the groups. We will now add a filter to specify a specific order of these groups, the order of our scenario!

@@@@  First put the "Secondary dimension" on "Keyword (Or Source/Medium) Path". By doing this we will get the same information overview about the keywords used.

Set the Secondary Dimension on Keyword (Or Source/Medium) Path

Click on "advanced" next to the search bar to enter a filter.

Click advanced.

Enter the following three filters:

• [Include] [Branded vs Non Branded Path] [Containing] [2 Branded CPC Searches]
• [Include] [Branded vs Non Branded Path] [Begins with] [4 Non Branded CPC Searches]
• [Exclude] [Branded vs Non Branded Path] [Ends with] [4 Non Branded CPC Searches]

Enter these three filters.

So now we only keep the data of the people who have clicked at least one time on a branded ppc ad, who have clicked first on a non-branded ppc ad and not ended by clicking a non-branded ppc ad (these conversions are included in the non-branded campaign according to the last-click principle). Click on Apply.

Results

Voilà! The searched report is now shown.

The sought end report.

It is very useful to just scroll through this list. You can see very well how our scenario goes and what your customers do in between these operations.

The scenario like we thought about in the first place.

On top of this page you can see the total profit of this report: "€9239,60,-". So this is the profit of the branded campaign who is supposed to be in the non-branded campaign. Of course, the costs we made stay the same. The newest data are:

• Non-Branded Keywords:
  • Costs: €14.353,-
  • Profit: €109.422,-
• Branded Keywords:
  • Costs: €6.017,-
  • Profit: €250.638,-

Conclusions

The corrected ROI’s are 7.62 and 41.66 for the Non-Branded and Branded Keywords. This is a deviation of 9% and 3.5% of the original values. If this is an decisive amount, is to discuss. In this case it is not. I would keep advertising on branded keywords for this company. However, if your target ROI for a branded campaign is reached before handed, but marginal, this little difference can be an extra reason to stop advertising on the own company’s name. In this case keep an eye on the profit and calculate, with the help of the total ad costs, if you are making progress or not. What conclusions could you get out of this? Did you come across any other interesting applications?