How fintech company Finom expanded into new markets

Finom, a B2B financial service startup which provides online financial services for SMEs, freelancers and the self-employed in Europe, has a proven answer to that question. Finom partnered with Lira on digital marketing to test performance channels, including Google Ads. 

Headquartered in the Netherlands, Finom started as a global player from day one. After a successful launch of invoicing MVP in Italy, they started to expand rapidly into other European markets with other banking services. 

Their next target market was France, and then Germany, where they faced severe competition and thus a high cost of customer acquisition. The challenge was clear: how to obtain a fair share of such a mature market as the German one, yet stay profitable?

Based on the experience with the launch in Italy, we started German expansion with the following campaigns:

1. Search, targeting banking service keywords [Geschäftskonto, girokonto für selbständige, bankkonto umsonst / Business account, current account for self-employed, bank account for free]
2. GDN, targeting custom Intent audience with the same keywords;
3. GDN, targeting web-site visitors, excluding already onboarded clients.

All the campaigns were divided into 2 buckets with different CPL targets: 1) targeting users that haven’t interacted with Finom at all; 2) ones that are familiar with the business (mainly by being on the web-site), but haven’t converted yet. With time we added more KWs to the second (remarketing) bucket, including those shown in Recommendations tab in Google Ads account.

At the beginning we tested broader (long-tail) KWs related to opening a bank account for a business, trying different variants. The most effective KWs turned out to be those containing ‘business account’ [Geschäftskonto]. Other resultive KWs were ones implying an intent to open one, e.g. ‘current account for self-employed’, ‘bank account for free’, ‘current account opening’, ‘account online banking’, etc. we didn’t use any competitors KW at first, but that changed at the scaling phase.

Being a banking service provider, Finom measures their success by onboarded clients – those that have passed all the verification steps and subscribed and now enjoy the online banking and invoicing services to the full. The first month of testing revealed a way too high CPL, roughly 3x of the target. Most dramatically, the conversion rate from a lead to an onboarded client was about 7% (it did go up to 40% later though). The margins were at risk and something had to be changed.

To scale up effectively and increase low-funnel conversions, the goals for the campaigns and bidding strategies had to be changed from Sign-ups leads to Onboarded customers. This would help to hedge against poor quality leads and high CPA. 

It was vital to hit about 50 onboardings per month to make sure targetCPA bidding strategy would work effectively. After setting up low-funnel conversions in the Google Ads account, it took a couple of months to reach this threshold, accumulating historical data. Once the volume was there, the low-funnel conversions were set for bidding strategies as a primary goal. That was the first step towards scaling.

At this point we partnered with Google to hit the ambitious goal: increase the amount of onboarded customers by 10x from the German market, which still remained quite a challenge.

The idea behind using Broad Match was that there were lots of generic queries left untouched, so it would help to increase the keyword coverage. It would also save time on brainstorming and testing new keyword ideas for Search campaigns.

The Recommendation tab in the Google Ads account was signaling (based on auction-based modeling) that there would be an uplift in traffic and leads from adopting Broad Match.

We held an A/B test through Experiments for 14 days. The traffic was split in 50:50 proportion. Campaign A had about 60 keywords in the broad match modifier, while Campaign B had the same keywords, but top-6 of them (by clicks and CR) were moved to broad match. The CPA target and the bidding strategy (tCPA) remained the same. 

By the end of the test Campaign B showed 37% more conversions, -28% in cost per conversion and +2pp in CR. Seeing such a positive result, we moved tested keywords to Broad Match permanently. Slowly, but surely more keywords were moved to Broad Match, allowing Finom to be on track with their ambitious target for the German market. By the end of the reporting period it was clear that about 85% of the leads came from Broad Match KWs.

To move further, Finom had to step into the competitors’ domain. We added branded KWs of the most active direct competitors mixing them with intent queries, but this time Finom tailored creatives and landing pages. 

After researching the competitive landscape, Finom redefined offers and changed creatives and landing pages. They moved their focus from general propositions to their strengths or unique features that competitors were missing. Landing pages were re-designed in such a way that would make Finom stand out and seem more attractive than a given competitor – they contained comparison tables highlighting the differences between products. 

The first screen of landing page in the Competitors campaign before

After

The approach paid off – Finom started to get conversions from the competitor KWs on a regular basis, sometimes at a lower CPL than from the hottest purely intent-based queries. Overall, the cost per onboarded customer decreased more than 2x.

Now we test new approaches and new beta products, like value-based bidding and Performance Max campaigns.

• Set-up a strong measurement and gather data to leverage the optimizing capabilities of Google Ads, – hedge against poor leads; 
• Don’t be afraid to be bold, – test new approaches to make sure it works for you;
• There are lots of fish in the ocean, – sometimes the ocean may surprise you!

Aleksandr Akhmadyrshin

Project manager, Lira Agency

Xenia Volokitina

Performance Marketing Specialist, Lira Agency

Tigran Chakhoian

Associate Account Strategist, Google