← Selected work

Case study · App launch, Europe

EVDC: 140,779 clicks and 910 installs in 33 days, from one price-comparison idea.

An EV-charging app launched cold in Italy, Spain and Germany. I made every ad. One video, "Same charger. Different price.", took 46% of the spend and won in both phases. Every number below is from the ad account, and the limits are stated.

Client
EVDC
Sector
EV charging app
Channel
Meta ads
Markets
Italy, Spain, Germany
My role
Angles, scripts, video, buying, analysis
Period
Oct 15 – Nov 16, 2026
Measure
Link clicks, then app installs
Source
Meta Marketing API, Nov 17, 2026. All figures CAD.
140,779link clicks, Phase 1
CA$0.08cost per link click
910app installs, Phase 2
CA$1.55cost per install, Italy
6.13%CTR on the best ad
CA$13,673total spend

Where it started: no audience, no proof, three markets.

EVDC had an app and no Meta history. On Oct 15 the account spent CA$15. By Oct 16 it was spending about CA$350 a day, and I was testing angles in Italian and Spanish at the same time.

01

Phase 1: traffic (Oct 15 – Nov 11)

Optimised for link clicks in Italy and Spain. CA$11,337 bought 3.22M impressions and 140,779 clicks. The job was to find which angle people would stop for, cheaply.

02

Phase 2: app installs (Nov 12 – now)

Switched the objective to installs in Italy and Germany, iOS and Android. The Phase 1 winner went in first. CA$2,336 bought 910 installs.

The timeline: four segments, one clear peak.

Daily spend over the 33 days, with the install phase in green. The table splits the run into the segments where performance changed.

■ traffic phase   ■ install phaseDaily spend, Oct 15 → Nov 16, 2026 (CAD)
Performance by time segment
SegmentSpendLink clicksCost / clickInstallsCost / install
A · Oct 15–26CA$3,86961,475CA$0.063––
B · Oct 27–31CA$3,84140,538CA$0.095––
C · Nov 1–11CA$3,33035,845CA$0.093––
D1 · Nov 12–13CA$1,3466,455CA$0.209129CA$10.43
D2 · Nov 14–16CA$1,2864,661CA$0.276781CA$1.65
TotalCA$13,673148,974910CA$2.57*

Source: Meta Marketing API, account-level daily series, reconciled to lifetime spend of CA$13,672.96. Table clicks are account-level link clicks per day, so they include clicks on the install campaigns. The 140,779 headline is the traffic campaigns only. *Blended cost per install on the app campaigns (CA$2,336 ÷ 910).

Best cost-efficiency: segment A, at CA$0.063 a click, before I scaled spend. Best scale: segment B, where I roughly doubled daily spend to about CA$770 and clicks stayed under CA$0.10. Best install window: Nov 14–16, at CA$1.65 per install after a rough first two days at CA$10.43 while the algorithm learned.

The best ad: "Same charger. Different price."

IT_LP_VID03-Offer is an animated offer video for the Italian market. It spent CA$5,200, which is 46% of Phase 1, and delivered 41.7% of all link clicks.

  • Reach: 1.12M impressions, 58,660 link clicks.
  • CTR: 6.13% all clicks, 5.25% outbound, on a CA$5,200 budget. Not a small-sample fluke.
  • Attention: 159,913 ThruPlays, about 14% of impressions watched through.
  • Second life: in the install phase the same angle was the top Italy ad, with 83 iOS installs at CA$1.30 each.

The install-phase ad was re-uploaded, so it carries a different ad ID. It's the same creative and angle, not the same record in Meta.

Why it works.

01

The angle: price, not features

Every charging app says it's fast and everywhere. Nobody says what the same session costs. "Same charger. Different price." makes one claim a viewer can check in seconds, and it makes the app the cheaper way to use a charger they already know.

02

The pain point: paying more for the same electrons

Drivers see different prices at the same plug depending on the card or subscription they use. The adjacent angles I tested all push on the same bruise: no subscription, price lock, a fixed-rate offer.

03

The format: short animated video

Video beat static on CTR across the account. In Germany, animated offer ads beat UGC on cost per install by about 2.5x on iOS and 2.6x on Android. An animation can show a price gap in two seconds, and a person talking to camera can't.

What didn't work as well is as useful as what did. Pixar-style and people-led ads were fine in Italy at CA$1.43–1.68 per install, but nothing in the account matched the price comparison for scale.

The ads with the best metrics.

Best ads by metric
Ad / groupSpendResultNote
ES_VID04CA$6577.58% CTRHighest CTR in the account
IT_LP_VID03-OfferCA$5,2006.13% CTRHighest-volume ad; won both phases
Italy, installsCA$695448 installs · CA$1.55Cheapest market
Germany, installsCA$1,641462 installs · CA$3.55Animated CA$2.47 vs UGC CA$6.93

Source: Meta Marketing API, ad and ad-set level, lifetime to Nov 16, 2026. The highest-CTR ad is a small-spend ad. The volume winner is the one I'd trust.

What this doesn't prove

  • Clicks aren't conversions. Phase 1 optimised for link clicks. A cheap click says people were interested, not that they installed or paid.
  • Five days of installs. 910 installs in Nov 12–16 is early. The CA$1.65 window followed a CA$10.43 start, so I wouldn't quote either as a steady-state CPI.
  • No revenue yet. I don't have charging sessions or lifetime value, so there's no ROAS here and I won't pretend there is.
  • Small samples. Most Germany ad sets have tens of installs. Animated beating UGC is a strong signal, not a settled result.
  • No baseline. There's no prior agency or in-house account to compare against. The market and the product were new.
  • Currency. The ad account bills in CAD. Figures are as reported, not converted.

On a call I'll show the ad-level breakdown and what I'd test next.

Next.

See what your account is leaving on the table.

Twenty minutes. You leave with at least one thing worth changing, or I end the call early and give you the time back.