CASE STUDY · CLOUD INFRASTRUCTURE, PAID SEARCH

From Trial Signups to Paying Customers: How UpCloud Cut Acquisition Costs by 60% in 90 Days

How Somebody Digital rebuilt UpCloud’s paid search account from the structure up, consolidating eighteen fragmented campaigns into one system precise enough to turn trial signups into paying customers, not just clicks.

-60% CAC +250% conversion rate 90-day trial Cloud infrastructure

0 %

Increase in customer conversion rate

In 90 days

- 0 %

Reduction in acquisition cost

Across the 90-day trial

0 %

Over-achievement of volume targets

By Month 3

0 %

Search impression share

At the moments that mattered

THE SITUATION

Eighteen campaigns, one budget, and no signal strong enough to bid on.

UpCloud had the product, the brand, and the traffic. What it did not have was a paid search engine that turned that traffic into revenue. The in-house PPC operation was generating a steady stream of trial signups, and on paper the numbers looked healthy, but trials were not converting into paying customers at a rate that justified the spend.

Eighteen separate campaigns competed for budget, data, and attention. That fragmentation meant no single campaign had enough conversion signals to let smart bidding do its job. The algorithm was optimizing blindly, and acquisition costs had climbed into the thousands per customer.

The demand was there. The challenge was building a program sophisticated enough to capture it efficiently.

Equal-sized campaigns splitting one shared budget and signal

Fragmented Signal

Separate campaigns split one budget and diluted the data. No single campaign carried enough conversion signal for smart bidding to optimize on.

Attribution Lag

CRM delay

A reporting gap between trial signup and confirmed customer meant bidding was optimizing for signups, not revenue, and had no way to know the difference.

Unsustainable Cost

$000s

Per customer in acquisition cost, unsustainable for a growth-stage cloud provider with real margin pressure and ambitious targets.

WHAT WE DID

Fix the structure before you touch the bid.

Somebody Digital started where the data pointed: the account structure itself. Five phases, each building on the data density the last one created.

Audit · Demand mapping

Geographic performanceMapped

Competitor positioningMapped

Keyword intentSegmented

Account structure · Consolidation

18 campaigns Diluted data Blind bidding

Unified
StructureONE ARCHITECTURE

Attribution · Closing the gap

Trial signup + Historical offline data

Real-Time
SignalBEFORE CRM CONFIRMS

Testing · Bidding and messaging

Smart bidding testsSystematic

Ad copy variantsRefined

Landing page alignmentMatched to intent

Live · Continuous scaling

Live · budget reallocation

4.31%Conversion

-60%CAC

40Customers

Budget moved away from trial-heavy journeys and toward what was performing.

Research and Analysis

Find where the demand actually lives

Before touching a single campaign, Somebody Digital audited geographic performance, competitor positioning, and keyword intent, mapping not just what was being spent, but where high-intent, transactional demand lived, and where budget was leaking into traffic that would never convert.

Account Restructure

Eighteen campaigns become one

The 18-campaign sprawl was consolidated into a tighter, market-based architecture designed to concentrate data density where it mattered. Fewer campaigns meant more conversion signals per campaign, giving smart bidding the inputs it needed to optimize for actual customers, not just clicks.

Attribution Framework

Bridge the CRM reporting lag

UpCloud's CRM data had an inherent delay between a trial signup and a confirmed customer conversion. Left unaddressed, this would have kept pushing bidding toward the wrong goal. Somebody Digital built a manual attribution framework using historical offline data, optimizing for customer outcomes in near real-time, before CRM confirmation caught up.

Bidding and Messaging

Sharpen the signal, click by click

With the structural foundations in place, Somebody Digital ran systematic smart bidding experiments, testing strategies, refining ad copy, and aligning landing page messaging to the transactional intent of the queries being targeted. Every test sharpened the signal between click and conversion.

Continuous Scaling

Let the account reinforce itself

As conversion patterns became clear, budget was actively reallocated away from trial-heavy journeys and toward the customer acquisition paths that were performing. The account became self-reinforcing: better data, better bidding, better outcomes.

THE RESULTS

A paid search program that became unrecognizable in ninety days.

Not from more spend. From a cleaner path between click and customer.

90-DAY TRIAL PERIOD

The headline shift

Customer conversion rate+250%

↑ Reaching 4.31%

Customer acquisition cost-60%

↓ Lower is better

Search impression share74%

↑ At the moments that mattered

SAME QUERIES, SHARPER TARGETING

Click-through rate

4.85%

Before

7.04%

After

Sharper targeting and rewritten ad copy lifted CTR without adding a single dollar of spend.

MONTH 3 — VOLUME TARGET

40 customers delivered

Minimum threshold20

Actually delivered40

Twice the minimum threshold, proof that paid search could function as a reliable acquisition channel, not just a top-of-funnel awareness tool.

VS. THE MINIMUM THRESHOLD

100% over-achievement

0 %

By Month 3, UpCloud’s paid search program was not just hitting its target, it was doubling it, on a channel the business had previously treated as a cost center.

In the short time we've been working with Somebody Digital, we've been incredibly impressed with how rapidly we've seen results. Most importantly, their use of experimentation and data-informed decision making has been invaluable in helping UpCloud see significant improvements in our media efficiency and toward our ambitious growth targets.

Manuel Heilmann

CCO, UpCloud

WHAT THIS PROVES

Not more activity, but a framework that made it count.

UpCloud did not need a bigger budget. It needed a structure precise enough for the budget it already had to work. Once every campaign carried enough signal to bid on, the algorithm did the rest.

FROM FRAGMENTED SPEND

Concentrated signal

Eighteen campaigns became one architecture, giving smart bidding enough data density to actually work.

FROM TRIAL SIGNUPS

Confirmed customers

A manual attribution bridge closed the CRM reporting gap, so bidding optimized for revenue, not vanity signups.

FROM STATIC BUDGETS

Self-reinforcing scaling

As conversion patterns sharpened, budget followed performance in real time, compounding the gains.

Somebody Digital’s work with UpCloud shows what happens when account structure, attribution, and budget allocation are treated as one system instead of three separate problems.

LET'S TALK

Could your commercial engine work like this?

Most clients are operating against a distorted picture and don’t know it yet. The first step is mapping what yours is hiding.