Lean Startup, day 12: How to Find a Product Name
Previously in Lean Startup Adventure, we explained how we wanted to maximize learning through a demo. The content of this demo is a landing page, hosted somewhere on the Internet. This landing page is an experiment in the Lean Startup sense, yet it needs to look polished enough to convince potential customers. And the absolute minimum for a landing page is a product name.
The final product name will probably change later in the process. At that stage, we just want a good enough name, chosen with limited budget. We decide to invest at most one hour with four marmelab collaborators in a brainstorming session. Nothing fancy: just brainpower, post-it notes, and a whiteboard.
Preparation
I gather 2 developers, a project manager, and an intern, into a closed room. It's important that they are from various backgrounds. I set up one clear objective: finding a name for a new web product to be distributed as SaaS. I list the constraints: it has to be an international name, available as a domain name, free of competitors, short, and it must sound good at least in English and French. The room has a paperboard, each participant has loads of post-its and black markers.
We run different brainstorming waves: Production, Pooling, Sorting and Pruning.
Production (10 minutes)
This individual exercise happens in silence. Each participant writes as many names as possible on post-it notes, one name per note. They focus on quantity, not quality - we'll remove bad names later.
To help bootstrap the production, I write down 3 lexical themes on the paperboard:
- cooking & gastronomy,
- sports & outdoor,
- colors
Pooling (10 minutes)
Each participant then goes to the paperboard to stick their notes close to the theme they refer to. While a participant stands up, other participants can think of new names and fill more post-its, inspired by what was reported on the board, using analogies or phonetic proximity.
The paperboard is full. Total contributions : 62 names. Not bad! These product names look like:
Data Chief | Tasty | Joli Josette | Datafriend |
Data Cooker | Passoire | Data Plant | Varapid |
Blue Cerise | Datacumoir | Data Plum | Datawind |
AirBlue | Marcel Pepper | BlueCider | Rosebud |
Marcell | Silver Salt | Concube | Bleuday |
Dataseed | DoDoney | Data Rider | DataPump |
Monsieur Jean | Peperandu | Waterboard | Misocell |
Recell | Emancell | Water Data | Data Prune |
Baby Blue | Amies Données | Swagetti | Rizoti |
Datatouille | Red John | Greencell | Epicell |
SaltCell | Pourpre Ciel | REST-aurant | BlueCurl |
Pasta Rosa | Filtra | Marcel Poivre | |
Paeli | AppleSel | Pomme Granade | |
Data Slicer | Parcell | Blue Sleigh | |
Swimcell | DataPulp | API Squash |
If you know the marmelab team, you won't be surprised that most of these names are puns, too. Unfortunately, the best ones only make sense in French.
Sorting and Pruning (10 minutes)
In a collaborative and chaotic way, everyone stands close to the paperboard and starts moving notes to form clusters. Names that are objectively bad for any reason (contained another brand name, too long, not pronounceable) are quickly moved to the trash.
We also use a voting system to eliminate the names with the least adhesion: each participant can draw 10 dots, one dot per note. Notes without any dot also end up in the trash.
Pretty soon we have reduced the number of possible names to 24, grouped into 4 clusters.
Availability Checking (10 minutes)
We start checking the availability of the related domain names using a registrar search engine, each participant checking the names of a given cluster in parallel.
Only 6 of our 24 names are available... and not the ones we prefer. Some names are parked (i.e. already taken but available for sell at a hefty price). At this stage, we don't want to invest more than a few dozen euros on the name, so we dismiss the promising names that are for sale for several thousand dollars.
Final Choice (10 minutes)
We gather the rest of the team to discuss the final 6 names openly.
People who didn't participate in the first place bring interesting feedbacks, and dismiss a few names for reasons we hadn't thought of. New names also come up during this phase. We try asking every collaborator to sort the 6 notes from top to bottom, but there is no common pattern. So we ended up using the same voting system as in phase 2.
Final choice: misocell!
Misocell doesn’t mean anything, and we’re fine with that. But some of us can hear a meaning in it:
- misocell sounds (in French at least) like Isosceles. It makes it easy to remember: “like Isosceles with an m”
- It contains the word “cell”, which is in the lexicography of spreadsheets (one of our reference ecosystems)
- It sounded a bit like “Excel”
- Ancient Greek speakers understand the “miso-” prefix as “hate” (as in misogynist, a person who hates women). “miso-cell”, the person who hates cells. Still in phase with our product.
- I’m a vegetarian, and I happen to eat miso soup quite often.
Buying the domain name (10 minutes, 20$)
This is easy and takes more time to describe than to actually do. We use the Gandi service to book the domain name for various tlds (.io
, but also .org
, .net
, .com
, .info
), and a credit card to pay for the 20$ of the total.
A few minutes later, we receive confirmation from Gandi, with the technical instructions to use this domain name for our demo. We then choose the main tld we'll use for the product: it will be misocell.io (because the .io tld is already associated with many good SaaS products).
Next Steps
Misocell is a good enough name for what we want to do with it. We know we may change it in the future, so discovering that it means an insult in Swahili wouldn’t be a big deal. We are pretty satisfied with the return on investment: only one hour to come up with this name, that's cheap.
In the next post, we will describe how we realized a teaser video of the misocell product without actually developing the product. Stay tuned!
Credits:
Thumbnail picture: the eyes have it, by Vicki DeLoach