Scaling things that don't scale

Scaling things that dont scale

Olivier PailhèsLast updated on February 23, 2023
5 min

Ready to build better conversations?

Simple to set up. Easy to use. Powerful integrations.

Get free access

Ready to build better conversations?

Simple to set up. Easy to use. Powerful integrations.

Get free access

I’ve always loved doing things that don’t scale. Even before knowing who Paul Graham is or what he wrote. It’s about putting a little bit of craftsmanship into an impersonal world and an impersonal internet overall.

That proved very useful when we started Aircall. Obviously the product wasn’t working so well, so we had to  make up for the bugs with personal, one-to-one dedication. And we learned a lot from that.

But then we started having more and more customers. Recently, I realized we had customers I wasn’t aware of, names I discovered when Slack notified me with an upgrade or an invoice, while a few months before I could give you the first name of every single user we had, their configuration and their typical usage.

At the same moment – and that’s not a coincidence – I understood we couldn’t sustain our growth rate without changing things, and finally doing things that scale!

The mistake of using “scaling softwares”

As a CEO, you look at the 3 buckets and try to find ways to scale that. The most obvious one that hits you is using better software applications.

  • Instead of building a list of prospects manually, we started using BuiltWith.com and crawling directories with KimonoLabs.com to build email databases. We plugged PersistIQ to automate email outreach and follow-up.

  • We also implemented automated onboarding emails during the trial period as well as post-subscription using intercom.io.

Of course, we tried to be as personalized as possible, filtering leads that would actually benefit from our value proposition, and triggering nurturing emails based on customer behavior.

On top of that, I realized we had a very low inbound acquisition flow, so we worked on SEO, on Content Marketing, and overall we increased our web presence on any platform that our target customers could possibly visit.

It actually worked, but the quality of our prospects and the relationship with our customers suffered from it. Plus, I was feeling uncomfortable acting as a robot.

Scale resources, not customer interactions

Luckily enough, we got funded at this moment. When you’ve raised, you then have to decide how you’re going to spend it. We chose to scale resources before anything else.

We hired a few sales rep to grow our Bus Dev team. For lead generation, we started using outsourced lead generation services from Upwork rather than scraping tools or lead generation softwares. You see, in the end, I want individuals dealing with a specific geography or customer group just like I handled Aircall’s first customers.

That’s expensive, and above all time consuming. You can’t expect the same level of knowledge and involvement from new hires as what you put in your own company upfront. But at least, with individuals, you can convey a culture of sweating for customers rather than letting a software solution deal with them.

Of course, we kept automated nurturing emails, but there a few things that we put back in place

  • Immediate phone calls when a customer signs up. This tip comes from a great discussion we had with Steli Efti from Close.io. Even today, as a quite grown-up startup, they keep on calling all of their new signups. That’s something we’ll keep forever.

  • 1-on-1 demos. Sure, our product is self-service. But making a personalized demo and helping the customer understand what she can get from the product is priceless. Of course it means dedicated resources – worse, trained and skilled resources. But it’s definitely worth it!

  • Manual follow-up based on judgement. Should a customer poorly rate a call ? We’ll send him or her personalized email with recommendations and follow-up questions –not always, but quite often.

Another  smart and scalable idea is NPS system Mention implemented to reduce customer churn. It actually goes beyond the simple question “How likely is it that you would recommend our service to a friend or colleague?” The score leads to a more specific question, which then triggers an auto-generated email based on the answer.

NPS Mention

For instance, if the answer is ‘I’m not ready to upgrade yet’, the user is then  asked to share his reasons in exchange for two extra weeks of free trials.

Even though, Mention’s team recalls that customer interactions cannot be 100% automatized. Their NPS process is merely an excuse to get their customers to talk to them and tell them more about their needs and wishes. They automatize individual interactions to the point where they can actually talk to them.

Doing unscalable CEO things that make the company scale

Once you’re scaling resources, you shift to a rather managerial role. Essentially implementing continuous improvement processes and making sure the team is great and happy.

But there are a few things you can do as a CEO to keep your craftsman’s roots :

  • If, as we currently are, you’re moving up from small customers to larger accounts, you can start over the hand-made customer acquisition, activation and retention process. At this point, flagship customers can really boost growth . Plus, here come the demos and 1-on-1 interactions!

  • Making pilot sales effort to new customer targets. Nobody else than you knows best how the product can used and by whom, beyond primary target. I’m trying to identify customers whose usage completely differs from our existing customer base. Most of our customers are web businesses or startups, and we’re already scaling on this target with the team. What if we could broaden our reach to the traditional Real Estate sector? Or Travel agencies? Can I find a few target customers that validate a use case, maybe with a slightly different product? Essentially, it’s about paving the way from the Innovators to the early adopter group, or to the early adopter group to the early majority.

(Yeah you probably know this chart)

Product adoption curve

  • Level 2 support. After dealing with customer issues and concerns for the last 18 months, I have piled up quite an experience on support. We actually don’t have a structured “Level 1 / Level 2” support, but I try to keep an eye on all customer questions or issues to jump in when needed. If nothing else, it ensures that I keep in touch with the product inside-out.

My $0.02. Hope it can be of any help or food for thoughts to some of you.

As this is a bit of a cathartic exercise, I realize we’re still very far from doing things that scale… 😉


Published on July 16, 2015.

Ready to build better conversations?

Aircall runs on the device you're using right now.