PerformanceININside Performance Marketing
JOIN THE PERFORMANCEIN Top 50 Industry Players from 2017
How to Ensure Synergy Between Your Social & Search Performance Channels

How to Ensure Synergy Between Your Social & Search Performance Channels


Since its inception, Social advertising has too often been viewed as a separate entity to other, more traditional, digital channels such as Paid Search. However, an integrated cross-channel strategy will enhance performance and assess the true value of all your activity, as well as improving the customer experience at all touchpoints.


One of the key challenges facing Paid Social practitioners is the lack of transparency surrounding sales attribution on Facebook. It’s all very well and good saying, “this campaign drove £X in revenue with a ROAS of £15”, but when analysis isn’t undertaken to de-dupe against other channels whether that be Paid Search or Display, your optimisation decisions are missing key information.


With third-party view tracking still unavailable on Facebook Custom Audiences, we are fairly limited in our ability to achieve true attribution. As a result, our preferred method is to utilise post-click tracking via solutions such as Google Analytics or DCM. This can then be de-duped against all your digital channels.

With Paid Social advertising, many assume it is difficult to see returns when using post-click tracking, however; we have seen promising results from direct-response campaigns. In a recent campaign, we utilised post-click de-duping between Paid Search, Paid Social & Display and saw 70% of Facebook-attributed post-click sales de-duped on a last-click mode (see graph below).l. However, despite this, we still saw a ROAS of over 400%.


It is essential to overlay this with insights from Facebook’s attribution solution. Whilst post-view can’t be effectively de-duped, there is an added value to actions that were influenced by viewing a Facebook ad. This is particularly the case with branding campaigns which might centre around video or canvas views as opposed to driving traffic to a website.

Therefore, it is important to adopt a flexible and multi-layered approach to attribution modelling when running Facebook activity alongside other channels. A de-duped post-click gives us a clearer view of true ROAS, allowing us to make optimisation decisions such as distributing budgets across channels and evaluating the performance of various targeting audiences. However, this would not be a complete view without also incorporating the value of post-view. With both of these, we can optimise based on facts, rather than opinion and conjecture.

Align campaign goals

So, you’ve defined your rules for attribution - the next step is to set your campaign goals across Paid Search and Social. The key here is to ensure your campaigns align and complement each other. Make sure you play to each channel’s strengths. Paid Search is extremely adept at picking off the low-hanging fruit: customers who have declared their intent and are close to making a purchase, but may be unsure about where to spend their money. Paid Social on the other hand, is a different beast. Most Facebook or Instagram users are looking to be entertained, informed and to generally pass some time.

Intelligently targeted ads with engaging creative can satisfy these needs whilst exposing new customers to your brand or re-engaging with existing customers. Consequently, advertising on both is an important part of a full-funnel approach.

Cross-channel re-marketing

Intelligent use of audiences in cross-channel remarketing allows you to utilise data from one channel for increased granularity on the other. The key for this is to make sure you tag all of your campaigns across search and social with unique parameters, the more granular the better. With these tags, audience segments can be utilised across all channels, taking into account the differing value and behaviour of customers at different stages of the purchasing funnel.

For instance, if you’re having a sale, target non-converters who clicked through a full price Google Shopping ad with discount focussed Social ads. For those who interacted with Instagram Story ads, which are historically low converting, re-target them with product carousels on Facebook and more aggressive bids on Google Shopping. During sale periods, create an audience of all those that clicked through your sale related keywords and converted. This discount motivated audience can then be retargeted across all your channels in future sale periods with increased bids and tailored ad copy. Ultimately, the more granular the tags, the wider the range of targeting options at your disposal down the line.

Additionally, these audiences give insight into the lifetime value generated by your prospecting campaigns. Audiences created from these interactions unveil how brand-aware customers behave in the future as they come back to convert through Paid Search or Social remarketing campaigns. The increased understanding of how prospecting impacts results will allow for more efficient budget allocation.

Consistent messaging

Whilst messaging best practice varies across Social advertising and Paid Search due to elements such as character limits, audiences, ad appearance and more, there remains a need to maintain a consistent brand image across all channels. Moreover, properly applied sequential messaging, using the audiences from the cross-channel remarketing lists, helps create a coherent customer journey from first to last touch point.

Beyond ensuring consistent ad copy themes, it is vital that testing results influence copy for each channel. For example, if a headline has attracted a high CTR in your Social ads, test it as a “Headline 2” in some of your Search campaigns. What works on one channel won’t necessarily work on the other, but these tests can still offer valuable insights into the behaviours of in-market audiences across both channels.

Product-level data-sharing

Google Shopping gives us a wealth of data when it comes to individual product performance, both through search volume and sales. Utilising this data at a product level is essential for a truly granular Google Shopping approach.

However, it can also influence strategy for Dynamic Product Ads on Facebook. By inputting Google Shopping performance data into the Facebook feed, product sets can be created that dynamically reflect performance, rather than static, arbitrary divisions based on category or price. As a result, bidding can be tiered according to the true value of each product set.


To ensure synergy between your Social and Search advertising, take the following key steps:

  • Define a clear attribution model – what value do you place on post-view? What attribution window do you want to use? How will you de-dupe your post-click sales? Confusion at the beginning will only lead to inefficiencies down the line.
  • Tag up everything – tagging all campaigns across both channels allows granular cross-channel remarketing, efficient budget allocation, relevant sequential messaging and more.
  • Share data as well as audiences – don’t be selfish with good performance, whether its top performing products or engaging ad copy, share it across your channels.    

Combining these steps with the integration of new developments such as affinity audiences in AdWords, third-party post-view tracking and more will engender Social and Paid Search channels that represent two complementary pieces of the puzzles, rather than separate entities.

Continue the conversation

Have something to say about this article? Comment above or directly on Facebook, Twitter or our LinkedIn Group.

Jack Carr

Jack Carr

Senior Analyst at NMPi. Jack graduated from Durham University in June 2015 having studied Economics, Spanish & Business. After a few months of soul-searching, woodworking & part-time jobs ranging from selling gourmet marshmallows to location scouting for educational camps, he joined the Performance team at NMPi in late March 2016. Since then, he's worked across a multitude of verticals delivering high performing Paid Search and Paid Social campaigns, predominantly on a pure performance model.

Read more from Jack

You may also like…