Case studies / Stackly
Lifted Stackly’s ROAS from 1.6× to 3.9× in one quarter.
Stackly is a B2B software company selling workflow tools to operations teams in the UK and North America.
SaaS
Paid search
Conversion optimisation
Case studies / Stackly
Lifted Stackly’s ROAS from 1.6× to 3.9× in one quarter.
Stackly is a B2B software company selling workflow tools to operations teams in the UK and North America.
SaaS
Paid search
Conversion optimisation
3.9×
ROAS
53%
Lower CPA
54%
Pipeline growth
3.9×
ROAS
53%
Lower CPA
54%
Pipeline growth
3.9×
ROAS
53%
Lower CPA
54%
Pipeline growth
About
Stackly grew on a free trial that anyone could start in a few minutes, with a sales team picking up the larger accounts. Google Search carries most of its paid acquisition, and the free trial is the conversion every campaign is measured on.
About
Stackly grew on a free trial that anyone could start in a few minutes, with a sales team picking up the larger accounts. Google Search carries most of its paid acquisition, and the free trial is the conversion every campaign is measured on.
Challenge
Stackly’s cheapest trials were its worst ones.
Stackly’s paid search had been optimising towards trial starts since the account was built. Broad match keywords and one catch-all campaign pulled in support queries, job applicants and people comparing tools they had no budget to buy. Trials were plentiful and cheap, and very few of them became paying accounts.
Sales could not tell which campaigns were worth following up, and the account kept scaling towards the only signal it had been given.
Challenge
Stackly’s cheapest trials were its worst ones.
Stackly’s paid search had been optimising towards trial starts since the account was built. Broad match keywords and one catch-all campaign pulled in support queries, job applicants and people comparing tools they had no budget to buy. Trials were plentiful and cheap, and very few of them became paying accounts.
Sales could not tell which campaigns were worth following up, and the account kept scaling towards the only signal it had been given.
Approach
We changed what counted as a conversion.
We worked through Stackly’s search terms, landing pages and trial data to find what separated a trial that activated from one that never returned. Setting up a team workspace and inviting a colleague in the first week told us almost everything, so that became the event the campaigns bid towards.
Keywords were then grouped by how much the searcher already understood about the problem. Each group got its own landing page, its own match types and its own cost target.
Approach
We changed what counted as a conversion.
We worked through Stackly’s search terms, landing pages and trial data to find what separated a trial that activated from one that never returned. Setting up a team workspace and inviting a colleague in the first week told us almost everything, so that became the event the campaigns bid towards.
Keywords were then grouped by how much the searcher already understood about the problem. Each group got its own landing page, its own match types and its own cost target.
Execution
Intent decided where the money went.
Search terms, trial activation data and sales feedback set the structure of the account and the ceiling on what each campaign could spend.
Execution
Intent decided where the money went.
Search terms, trial activation data and sales feedback set the structure of the account and the ceiling on what each campaign could spend.
01
Read the search terms
Weeks 1–2
Twelve months of search queries were sorted by what the person was actually trying to do. Terms tied to support, careers and research became negatives across the account.
01
Read the search terms
Weeks 1–2
Twelve months of search queries were sorted by what the person was actually trying to do. Terms tied to support, careers and research became negatives across the account.
02
Rebuilt around intent
Weeks 3–6
Brand, competitor and problem-aware searches were split into separate campaigns, each with its own match types and budget. Cost targets were set against the value of the accounts each one produced.
02
Rebuilt around intent
Weeks 3–6
Brand, competitor and problem-aware searches were split into separate campaigns, each with its own match types and budget. Cost targets were set against the value of the accounts each one produced.
03
Matched the page to the query
On the landing pages
Competitor searches landed on a direct comparison, problem searches on a use-case page. Both dropped the credit card requirement and asked for a work email only.
03
Matched the page to the query
On the landing pages
Competitor searches landed on a direct comparison, problem searches on a use-case page. Both dropped the credit card requirement and asked for a work email only.
04
Bid on activation
Weeks 9–12
Trial activations were imported back into Google as the conversion, so bidding optimised for the trials that used the product. Spend moved behind the campaigns producing them.
04
Bid on activation
Weeks 9–12
Trial activations were imported back into Google as the conversion, so bidding optimised for the trials that used the product. Spend moved behind the campaigns producing them.
01
Read the search terms
Weeks 1–2
Twelve months of search queries were sorted by what the person was actually trying to do. Terms tied to support, careers and research became negatives across the account.
02
Rebuilt around intent
Weeks 3–6
Brand, competitor and problem-aware searches were split into separate campaigns, each with its own match types and budget. Cost targets were set against the value of the accounts each one produced.
03
Matched the page to the query
On the landing pages
Competitor searches landed on a direct comparison, problem searches on a use-case page. Both dropped the credit card requirement and asked for a work email only.
04
Bid on activation
Weeks 9–12
Trial activations were imported back into Google as the conversion, so bidding optimised for the trials that used the product. Spend moved behind the campaigns producing them.
Results
Stackly bought fewer trials and closed more of them.
Results
Stackly bought fewer trials and closed more of them.
3.9×
ROAS
53%
Lower CPA
54%
Pipeline growth
“
We were measuring the wrong thing and paying for it every month. Korva gave sales a reason to trust the trials coming through paid search.
Sarah Lin
Co-founder at Stackly
3.9×
ROAS
53%
Lower CPA
54%
Pipeline growth
“
We were measuring the wrong thing and paying for it every month. Korva gave sales a reason to trust the trials coming through paid search.
Sarah Lin
Co-founder at Stackly
3.9×
ROAS
53%
Lower CPA
54%
Pipeline growth
“
We were measuring the wrong thing and paying for it every month. Korva gave sales a reason to trust the trials coming through paid search.
Sarah Lin
Co-founder at Stackly
