Understand Your Conversions Better with Attribution Modeling

It is important for any business using Adwords to accurately track their conversions. This helps them optimise their Keywords and Ads to ensure that they get the best value for their marketing spend.

Defining Attribution Model

The process that determines how credit for sales and conversions is assigned to the various touch points across the marketing journey is Attribution Model.

Attribution Modelling enables a marketer with every bit of information related to what attracted to the customer, his social media interactions, and how much time it took him in the complete buying circle before purchase, and what were the factors that drove him towards the purchase decision etc.
This type of information proves helpful for the marketer to direct the marketing investments towards the best reciprocating channels and campaigns. And, it lets us organise our budgets in a better way, by not investing in less effective channels ensuring good ROI.

Adwords offer five different attribution models for Adwords Managers to consider:

  1. 1. Last Click
  2. 2. First Click
  3. 3. Linear
  4. 4. Time Decay
  5. 5. Position Based

For businesses that have more than 15000 clicks and a conversion action with at least 600 conversions within 30 days, Data-Driven attribution is also available. We will overlook this model as very few small or medium businesses have that sort of numbers available to them.

What data each attribution model offers?

There are five different Attribution Models that are widely used, but the challenge is in figuring out which model to go by. Since every buying cycle is different, attribution models aren’t a one size fits all solution.
We at Inbound Leads, regularly look at all attribution models depending on the customer’s business and the type of data we are trying to collect. These attribution models help us in better understanding of the role that each keyword and Ad play in generating the lead or sales for our customers.
Most advertisers measure the success of their campaign on the Last Click basis. This gives all the credit to the last keyword that generated the impression and the last ad that got clicked.

On the other hand, if you use the First Click attribution model, all the credit goes to the ad and the keyword that generated the click first.

On an average, customers click on your ads 1.5 times before they decide to contact you or buy from you. This is especially true for eCommerce websites and other b2c websites. Chances are that they would have used different search queries and clicked on a different creative each time they visit your website through your Pay Per Click ads.
Therefore, it is very important that each creative and keyword that helped bringing that customer to your website is given equal weightage when counting the conversion. This is where Linear Attribution model comes in play.

The Linear attribution model equally distributes the credit of conversion to all clicks across the customer journey. If a customer clicked on 2 different creatives of your campaign, it will distribute .50 value to each. Similarly, if they click on more than 2 creatives, the weightage is distributed accordingly.

Time Decay attribution model attributes higher weightage to the click that is closer in time. A click that is received 1 day before the conversion will get double the weightage than the click received 10 days ago. According to Google, the credit is distributed using a 7-day half-life.

Position Based attribution model gives 40% weightage each to the first clicked ad and the last clicked ads (and corresponding keywords). The rest 20% weightage is distributed across to other clicks in the customer’s journey.

Benefits of each attribution model

Marketers can get different type of data from each attribution model and use it to optimise their Adwords campaign. Each attribution model gives varied values to key touchpoints. Thus, it is very important to understand the behaviour of your target audience to determine the best attribution model for you.

Traditionally most marketers use the Last Click attribution model and in our opinion, it is over-valued. Putting too much focus on the last click takes away credit from the ad at the top of the funnel or other ads/keywords that would have assisted in constantly keeping your business in the front of the user.

  • Last Click: This model gives the maximum weightage to the click at the bottom of the conversion funnel. If you want the most efficient cost per conversion you should be using this model. However, this tends to give too much value to brand-terms & re-marketing. If you look at your data using this model, you will notice that a lot of the conversions are coming from your brand terms or very specific keywords. You might miss on the value offered by key-phrases that your customers are initially using to find your product or service.
  • First Click: This attribution model gives 100% credit to the Ad and Keyword that initially bought the customer to your website. This model helps you understand how the customer found you initially. The disadvantage of this model is that it tends to favour highly competitive key – phrases whereas people searching initially use more generic key – phrases.
  • Linear: This model assigns equal weightage to each touchpoint. This helps you understand the complete customer journey. However, the disadvantage of this attribution model is that it undermines key touchpoints while over valuing minor touch points.
  • Time Decay: In this model, most of the value is assigned to the last touch points while reducing the value of the earlier touchpoints. Like last click attribution model, this tends to put too much value on brand terms and re-marketing ads. This could be useful for companies with conservative growth strategy focussing on sales/conversions. This method will attribute incremental value with each click giving maximum value to the click closer to actual conversion.
  • Position Based: This attribution model assigns 40% of credit to the first and last touchpoints. The remaining 20% is distributed to the middle touchpoints. Although in short purchase cycle it might be ok, this becomes an issue when the purchase cycle is long and spreads across a few months. This attribution model would be ideal for companies with aggressive growth strategy as it focusses on bringing new customer into the sales funnel.

For most of the businesses the best possible attribution model will be Position Based as it focus on both Growth and Efficiency. It gives equal weightage to the search term that initially bought the visitor to the website (in most cases it is through competitive keywords) and the click that lead to the conversion. Although it could overlook the clicks in the middle of the funnel, it is many time the first and the last click that are most important touchpoints.

A Real World Example

John is a mortgage broker and his firm ABC Mortgages specialises in low cost home loans. A customer finds his site by performing each of the following searches “mortgage broker sydney”, “cheap home loans”, “low cost home loans” and “low doc home loans”. She fills out a lead form online after clicking the ad that appeared on searching for “low doc home loans”.
Using each of the attribution model, this is how the credit to each keyword will be allocated:

  • Last Click: In this attribution model, the last keyword “low doc home loans” will receive 100% credit for the conversion.
  • First Click: This attribution would give 100% credit to “mortgage broker Sydney” keyword and overlook all other keywords.
  • Linear: This attribution model will give equal credit for all 4 key phrases and thus each keyword will receive 25% credit.
  • Time decay: In this attribution model, “low doc home loans” will receive the most credit for conversion as it is searched closest to when the conversion happened. The least credit will go to “mortgage broker Sydney” as it leads to the earliest click and is furthest to the conversion.
  • Position based: This model will give 40% credit to both “mortgage broker sydney” and “low doc home loans”. The two keywords in the middle of the funnel “cheap home loans” and “low cost home loans” will only receive 10% credit each.

If you understand the home loan industry, you will know that this industry may have a long purchasing cycle and using each attribution model may have its own benefits and limitations.
A typical home loan buyer undertakes a lot of research before they make a purchasing decision or even talk to a mortgage provider. This means that the Last Click or the First Click attribution model will always have data skewed towards wrong key phrases and totally overlook the keywords that have assisted in generating the conversion. Similarly Position based or Time decay model tend to look over the key phrases that would have assisted the visitor during their research.
Linear model would possibly be the most suitable in this scenario as it gives equal weightage to the first and last click and to terms that would have helped ABC Mortgages remain in front of the user while generating the brand awareness and helping the visitor in their research.

What Model is Right for Me?

There are no straight forward right or wrong answers. The right attribution model depends on a lot of factors including:

  • Industry
  • Sales cycle – long or short
  • Your growth strategy
  • Target audience
  • Your Marketing objective – sales or branding
  • Your long-term objective & short-term goals

Just like the marketing strategy, the attribution model that is right for one business may not be the right one for your business. Consult Adwords experts or your partner agency before adapting any attribution model!

Leave a Reply

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