Search Lift

Measure incremental traffic from search engines, driven by Meta ads

ONE-PAGE SUMMARY

Search Lift

Getting started

1. Set up new UTM parameters (ideally one for

branded keywords and one for non-branded
keywords) for relevant Paid Search campaigns,
to allow for a custom conversion event to be
defined that tracks all site visits that come from
the search engine of interest

The same logic can also be applied to measure

organic search traffic

1. Create relevant custom conversions for Paid

Search site visits in Events Manager

2. Validate each new custom conversion by

comparing reported volumes in Events Manager
What is it? Why run a study? Requirements with your current web analytics tool (e.g. Google
Analytics)
Track website visits that come from Paid - Meta is a discovery platform. Discovery ads - Recommended: Conversions API, to allow for
Search in your conversion lift study and can drive consumers to search for keywords conversion lift testing 3. Set up a Conversion Lift study to measure
compare volumes across randomised test and (especially branded keywords) before incremental traffic from Paid Search, driven by
control groups, to identify search website converting - Recommended: Healthy Event Match Quality Meta ads
visits driven by Meta Ads score (>5) for site visit events
- A Search Lift can help show the more
immediate impact of Meta ads for advertisers - Ability to implement UTM parameters for
with a long purchase cycle or advertisers branded & non-branded Paid Search
testing tactics across the funnel (e.g. campaigns
upper-funnel campaigns) *E.g. Paid Search branded/non-branded keywords and organic search
Introducing Search Lift

Setting up a test
Agenda
Results

FAQs
01 Introducing Search Lift

4
People and brands intersect across surfaces
in organic and serendipitous ways across Meta technologies.

FEED STORIES SHOPS AD

MESSENGER GROUPS AR / VR REELS

5
People are uniquely engaged
across Meta technologies

People 80

% who shared or posted from platform

technologies

engage more 60

with brands 40

across Meta TikTok

20

technologies. Twitter

Snapchat

0 5 10 15 20 25 30

% who clicked on an ad in last month on platform

Source: GlobalWebIndex (online survey of 400,00+ internet users across 50 markets:
AE, AR, AT, AU, BE, BR, CA, CH, CL, CN, CO, CZ, DE, DK, EG, ES, FR, GB, GH, GR, HK, ID,
IN, IR, IS, IT, JP, KE, KR, MO, MX, MY, NG, NL, NO, NZ, PH, PL, PT, RO, RU, SA, SE, SG,
TH, TR, TW, US, VN, ZA), Q3–Q4 2022.
6
We don’t just create demand. We help convert it!

70%
58% DISCOVERY
of shoppers discover new
say Meta is the place products or brands on Meta
to build relationships technologies1
with brands2

LOYALTY CONSIDERATION

56%
46% saw an ad that drove
say Meta technologies consideration1
have helped them make PURCHASE
a purchase1

Sources: 1. “Discovery-Led Shopping Study” by GFK (Meta-commissioned online survey of 12,000+

1 billion
respondents ages 18–74 in the AU, BR, CA, DE, FR, GB, IN, ID, JP, KR, MX and US) Q2 2022; 2. “Global
Media Study” by Ipsos (Meta-commissioned online study of 48,517 people ages 13-64 who are internet people connect with businesses across
users, monthly mobile internet users, monthly video viewers and monthly messaging app users, across AU,
BR, CA, DE, FR, GB, IN, JP, KR, MX, SE, TH, TR and US), Dec 2021–Jan 2022. TR sample includes only
Messenger, Instagram and WhatsApp every week3
18-64; 3. Meta data, Feb 2022.
7
On average, attribution
undervalues Meta by

*
56%
Uncalibrated attribution On average, results attributed to Meta ads
in this study had to be calibrated

models don’t necessarily to incremental results by

**
measure the full value of 2.3x
advertising on Meta. Meta was undervalued
by uncalibrated attribution models in

***
59% of the studies

* calculated as: 1 - (1 / CM) where CM is the multiplier (2.3 on average)

** calculated as: CM = Incremental results / Attribution model results
*** Meta is considered undervalued when the multiplier (CM) is higher than 1.2
Results of research done by Meta in cooperation with EMEA based businesses that ran Meta
marketing campaigns. Based on 17 global Meta Conversion Lift studies ran between March 2022 -
September 2022. This analysis compares estimates from Meta ad experiments of Facebook and
Instagram marketing for clients to the value attributed to Facebook and Instagram
by their internal attribution models (values provided by the clients).

8
Our conversion lift methodology allows advertisers to
measure true value driven by their Meta campaigns

Test Group:

LIFT

Control Test

Control Group:

No Meta Ads

1. RANDOMIZE GROUPS 2. DELIVER ADS 3. OBSERVE CONVERSIONS 4. ANALYSE LIFT

9
Users who searched for your
LIFT
brand and visited your website,
because they had been

Search Lift exposed to your Meta ads

● Track website visits that come from Paid

Search campaigns in your conversion lift Control Test

study and compare volumes across test

and control groups, to identify search
website visits driven by Meta Ads

● Other conversions events (e.g. purchase,

add to cart) can also be tracked in the
same study, to allow for incremental
value to be measured across multiple
KPIs

10
• Helps you understand the true value of your

Why Search Lift? Meta ads in driving incremental search engine

traffic to your website and online assets.

• Meta is a discovery platform. Discovery ads can

drive consumers to search for keywords
(especially branded keywords) before
converting.

• Last-click attribution ignores the impact of

discovery ads in driving those last click
conversions.

11
Campaigns Across the Funnel

• When we optimize to events higher up the

Who is it for? funnel, it can be challenging to learn how these
activities impact bottom of the funnel results.
Search Lift can prove the value of Meta ads in
driving more people to search for their brand.

Long Purchase Cycle

• For products and services with a long purchase

cycle, conversion lifts might not reveal the full
impact of Meta ad exposure. Search Lift can
show the more immediate impact.

12
“Search Lift helped us uncover the value in our Brand Awareness
activity, which is always difficult to measure. We saw a significant
lift in branded search website visits caused by Meta Ads and
gained confidence in our investment”

Dror Pic
Head of eCommerce Digital Marketing
SodaStream

13
Luko Success Story
Luko is an insurtech company which is developing a
fully-integrated home care platform. It used Meta
technologies to promote its home insurance solutions.
Thanks to the integration of the Conversions API, Luko
were able to measure results with a conversion lift
study. Results showed a strong incremental impact at
all levels of the funnel. It also indicated that 18% of
search sessions were due to people being exposed to Success Story
the Meta campaign.

19% 9% 8%
incremental lift in incremental lift in incremental lift in
search sessions simulations run to subscriptions to a
for people calculate quote of home insurance
exposed to the a home insurance contract
Meta campaign

Dates: March 10th – May 10th, 2022

02 Setting up a test

15
Set-Up Requirements

● Recommended: Healthy Event Match Quality

score (>5) for all relevant site visit events , such
as PageView or ViewContent

● Ability to implement UTM parameters for Paid

Search campaigns

Set-Up Goal
● A custom conversion (created using UTM-based
rules) that successfully captures relevant site
visits from search

*Which site visit event that is relevant will depend on where your search campaigns direct people
when they click through. Speak with your account team if you need further details about event
definitions.
16
1 UTM Parameters

2a Custom Conversions

2b Optional Custom Conversions:

Organic Search Visits

Getting started 3 Validation

4 Conversion Lift Study

17
1 UTM Parameters

● Search Lift relies on the ability to track site visits

from search through UTM parameters. If your
business already have these implemented for UTM Example
the Paid Search campaigns you wish to track, (once implemented)

proceed to step 2. If not, new UTM parameters

needs to be created.
➢ Branded:
https://jaspersmarket.com/products/refill?utm_sou
rce=google&utm_medium=cpc&utm_campaign=jas
● You may wish to explore impact on different persmarket_campaign_x&utm_content=brand
types of campaigns, for example:
○ Branded keywords (recommended) ➢ Non-Branded:
https://jaspersmarket.com/products/refill?utm_sou
○ Non-branded keywords (recommended) rce=google&utm_medium=cpc&utm_campaign=jas
○ Google Shopping persmarket_campaign_y&utm_content=nonbrand

● In your search ads platform (e.g. Google Ads),

bundle all Paid Search campaigns with UTM
parameters based on the splits you’ve decided
to explore*

*If your campaign structure does not allow for paid search campaigns to be split into
branded/non-branded keywords, create one UTM parameter for all relevant and UTM-trackable paid
search campaigns instead

18
2a Custom Conversions

● Go to Events Manager to create a

Custom Conversion for branded Paid
Search site visits*

● Under Rules, enter the conditions

required to identify a Paid Search driven
site visit from the referral domain of
interest. For example**:
○ utm_source=google AND
○ utm_medium=cpc AND
○ utm_content=brand

● Repeat above steps for a non-branded

Paid Search site visits
○ utm_source=google AND
○ utm_medium=cpc AND
○ utm_content=nonbrand
* Or the campaign splits you’ve decided to explore
** Exact rules will depend on the UTM logic and naming taxonomy your business uses to track relevant Paid Note: Illustrative example only**
Search campaigns 19
2b Optional Custom Conversions:
Organic Search Visits

● Once there is no unique click identifier or

UTM in the URL and the referrer came
via a search engine (e.g. Google or Bing),
it is being categorized as organic search
traffic

● Go to Events Manager to create a

Custom Conversion for organic search
site visits, by specifying the referral
domain of interest (e.g. google) and a
rule for the URL not containing UTMs or
relevant unique click identifiers*

*E.g. GCLID for Google Ads/Google Search Ads 360, DCLID for Google Campaign Manager,
MSCLKID for Bing Ads

Note: Illustrative example only*

20
3 Validation

● Wait a few days before validating each new custom

conversion by comparing reported volumes in Events
Manager with your current web analytics tool (e.g.
Google Analytics), to make sure the events are
configured properly.

● Volumes may differ across sources by up to 20%. A

larger discrepancy will however reflect an event
configuration problem.

21
4 Conversion Lift Study

● Ask your Meta Client Solutions Manager or Marketing Science Partner to setup a Conversion Lift
Study to measure the Search Lift.

● Use Branded/Non-Branded/Organic visits objectives.

● You can also create a Total Search objective (organic search + paid branded search + paid non-
branded search).

● Test duration and size of the control group (“holdout”) will be based on a power calculations to
make sure you have well powered test.

● The test will entail a control group (“holdout”), which will not see Meta ads during the test. As part
of the power calculation, the size of the holdout will be determined. Usually it is between 5% and
15% of the target audience.

● The study activation and the holdout group should not affect your performance on Meta.

22
Set-Up Flow Overview
Advertiser:
Ad Ops Team

Search Advertiser: Meta:

Campaigns Developer or CSM / MSP
Set Up Implementation Team

UTM Conversion
Custom Lift Study
Parameters Conversions Set Up

Meta ad is served Paid Search ad is served

Randomized
User Test & Control
Journey Groups
Search by User Click by User Website Visit by User

(Randomized Control Groups)

Meta Pixel / Custom Conversion

Data CAPI Event Conversions Lift Study
Collection Fires Populated Populated
PageView or Data is
Search Lift
ViewContent Fired Validated
03 Results

24
04 FAQs

25
1. Do we need to touch our 8 events configuration for
the study?
ANSWER:

No. The 8 events prioritization is for optimization, not

measurement.

1. Is it possible to conduct the study without

Conversions API implementation?
ANSWER:

No

1. Can we measure Meta impact on website traffic from

other channels?
ANSWER:

Yes. Based on UTM logic you can also suggest to

track affiliate and any other web channel (YouTube,
Pinterest etc.) traffic

1. Can we measure Meta impact on direct website

visits?
ANSWER:

Yes. Create a custom conversion for direct traffic

(see on the right)
Note: Illustrative example only
26
Thank You