When you set goals for your paid traffic contractor, think about how these goals affect your profits. Simply aiming to get more leads within a budget can sometimes lead to lower profits. Let me share three case studies from my own experience to show you better ways to set goals that can actually increase your profits.

 

Case study #1

These are the statistics from running campaigns for a year, with a five-digit sum of USD spent during this period.

A common approach:

• Pause campaign 3 because it has a higher CPL.
• Set an equal budget for campaigns 1, 2, and 4 as they have similar CPL.

Results:

• Campaign 3, which is highly profitable, is paused, resulting in a loss of profit.
• Campaign 1 has a budget that is three times lower than it could be (if you were to use the budget from campaigns 2 and 4, which perform worse) resulting in a loss of profit.

My approach

• Pause campaigns 2 and 4 as they have lower profitability
• Use the freed-up budget for campaigns 1 and 3 as they have much higher profitability

My result:

• 2x increase in profit (all campaigns had the same budget)

 

Case study #2

A common approach:

• Pause age clusters 25-44 years old because they have a higher CPL.

Result:

• A drastic increase in CPS leads to a decrease in the number of sales, resulting in a loss of profit

 

My approach

• In this case, I paused age clusters 45-64 years old because they have a high Cost-per-Sale.

My result

• CPS decreased by 33%, leading to a 50% increase in the number of sales, even though the CPL increased almost four times.

 

Case #3

A common approach:

• When you task a paid traffic contractor with generating more leads or lowering the cost per lead (CPL), they typically take several actions. These include increasing the conversion rate from click to lead, raising bids and budgets, adding more targeting options and traffic channels, and removing more expensive targeting options.

Result

• If you don’t track the quality of leads in an ad account, the built-in AI can’t distinguish between good and weak leads. Aiming to ‘get more leads’ or ‘decrease the CPL’ may result in attracting more poor leads, which would lower your profits.

My approach

• In this case, I’ve set up the transfer of lead quality information to Google Ads

My result

• Google’s built-in optimization algorithm changed the proportion of keywords, campaigns, and locations, increasing the share of quality leads by 40%. This means a potential 40% increase in sales for the same budget.

 

Notice how the structure of traffic that generates ‘just leads’ differs from the structure of ‘quality leads’ traffic. I didn’t change any settings after the switch; everything continued with the same setup: keywords, ads, locations, etc.

Google’s algorithm changed my bidding amounts for different days and times of the day,

it changed the proportion of the budget I spend across different ad groups,

locations,

and keywords.

These are the settings visible to a user. Many more signals are factored in under the hood.

 

Current experiment

Challenges of this project

Optimization towards purchases is not feasible due to:

• A significant delay between clicks and purchases, which in this project can range from 3 to 9 months. This delay substantially impacts AI optimization because the algorithm struggles to account for events that occur far in the future.
• Observing, for example, five high-ticket purchases from a targeting option such as a keyword does not necessarily mean I should increase bids, because the next five purchases might be of much lower value.

As a temporary solution, I’ve begun tracking the quality of leads in the ad account, which increases the proportion of high-quality leads. However, this doesn’t indicate whether the current cost per lead (CPL) meets the return on ad spend (ROAS) target, due to:

• A wide range of average order values across products (from $50 to $10,000 per purchase), making it inappropriate to assign a single estimated revenue-per-lead across all channels and targeting options.
• A broad variation in probable conversion rates and average order values across different user locations (over 50) and languages.
• A large volume of traffic where it’s unpredictable which product the user will be interested in, not to mention that different brands and models of the same product may have different margins.

My solution

I am now passing estimated Revenue-per-Lead data into the ad account shortly after leads are generated. This approach considers how much revenue leads with similar characteristics have generated in the past. The algorithm will learn from leads generated by all traffic sources over the last two years.

Result

Check back here in a few months to see the outcomes of this experiment.

 

Next steps

To make your ad spending more profitable:

1. Ensure your ad systems get information about the lead quality and estimated revenue from your CRM.

2. Set goals for your paid traffic contractor that aim to boost profits, not just get more or cheaper leads. Here are some better options:

• “more sales under a target ROAS”,
• “more sales under a target Cost-per-Sale”,
• “more leads, with the estimated Revenue-per-Lead exceeding the CPL by 150%”,
• and “more quality leads under a target CPL”.

3. If you’re considering hiring me to manage your paid traffic, first see if I’m a good fit for you. Then, feel free to contact me so I can assess how I might be able to help and determine the potential cost.