How to effectively streamline media spend

With the paid search landscape becoming ever more competitive, and accounts becoming much more sophisticated in their set-up, we wanted to explore how advertisers can continue to drive growth from mature paid search accounts. In the first part of this series, we will discuss how marketers can streamline media spend to prioritise the most valuable users.

In order to achieve this we’ll be exploring two methods:

  • Audience-first strategy (applicable to all Google Ads accounts)
  • Advanced Smart Bidding (stronger results but more reliance on technical set-up)
PPC blog keywords and conversions


Different people will be searching for the same keyword with very different intents. In this article we’ll be reviewing strategies that can help to overcome this challenge.

For example:

  • Different average order value: The term “Holidays in the UK” is ambiguous. The user behind this search could be looking for a £25/night camping weekend or they could be looking for a £600/night cottage with a hot tub.
  • Different lifetime value (LTV): The term “Tiles and grout” is ambiguous. The user behind this search could be a tradesman looking for local supplies for the next three months or they could be the average person doing a small touch-up.
  • Different products: The term “Business insurance” is ambiguous. The user behind this search could be a plumber looking for Public Liability Insurance or an IT contractor looking for Professional Indemnity Insurance.

When the user visits the website, their intent then becomes apparent BUT a cost has already been incurred getting them onto the website. To streamline our costs and to find efficiency, ideally we need to know their intent before we pay for their click. This is possible with a sophisticated audience strategy.

Audience-first strategy

If we take the “Business insurance” example and our offering is only applicable to tradesmen, we need to filter out a large portion of traffic; otherwise we run the risk of paying for irrelevant users (even though the keyword is highly relevant).

Step 1: insights 

Create a portfolio of your most valuable customers (Google Analytics (GA) works best as we have access to all affinity and in-market audiences by default). In our example here, we may find that our most valuable users are:

  • 80% of the time in-market for van related topics, such as insurance or rental
  • 90% of the time in-market for home & garden DIY
  • 70% of the time they are on their mobile
  • 80% of the time they are male
  • Other key information around time of day / locations / day of week / age / customer match etc.

Step 2: action 

Find those trends that really differentiate your valuable customers from irrelevant users, as shown above. Then consider creating a priority campaign that only targets those segments. Now when we serve an ad for “Business insurance”, we have a much higher probability of reaching the valuable customers, sharply reducing budget wastage. Leftover spend can then go to the other campaign as normal, or consider being proactive and reaching out to your audiences via formats such as Discovery and YouTube.

Step 3: automation

Finally, review, adjust and apply automated bid strategies to achieve the best results.

Audience first strategy

Advanced Smart Bidding

Smart Bidding, also known as automated bidding, is a highly effective way of improving performance. However, its success has a high reliance on data and technical set-up. For the bid strategy to only pay for the most valuable customers, we need to give it a very clear definition of what is “valuable”; this can be done in a few ways:


If we take the example “Holidays in the UK”, which could result in a £25/night or a £600/night purchase, optimising towards revenue is a must and happens to be the most straightforward method we’ll be demonstrating in this article.

In this example we’ll be wanting to employ a Target return on ad spend (ROAS) bid strategy (utilise the Auction Time Bidding feature where possible). Google Ads tags, GA

Transactions and Search Ads 360 (SA360) Floodlights can all capture revenue. Once the bid strategy is in place, the machine will begin to find the most effective users and only pay for those who are most likely to convert and also those with the highest average order value.

Additional data

Incorporating additional data can fuel the machine with context so the bid strategy doesn’t just maximise conversions, it prioritises media spend to those who are most likely to continue to engage with the brand in a positive and profitable way. Here are two ways to improve the data we feed into the bid strategy.

  • Floodlights – Floodlights in SA360 are advanced conversion tags. Rather than just recording a conversion, Floodlights have the ability to review the data fields populated by the user, allowing us to break the conversions into multiple categories depending on the input of the user and therefore enabling the bid strategy to optimise towards the most valuable.
    Find out more about using Floodlight tags here.
  • Business Data – Business Data in SA360 can feed the machine profit margins, return rates, capacity, availability and more, each one giving the bid strategy more information to find value.
    Find out more about using business events with bid strategies here.

Hopefully this article emphasises the importance of understanding our opening statements; “Keywords are not enough” and “Not all conversions are equal”. Each user journey is different, each website is different and the most effective approach always requires a tailored strategy. For more paid search solutions please get in touch.

by Conor Kingscott
15 April 2021



Related posts