Branding for Startups is BS. Stop Trusting Experts, Trust Data.

Yep  —  we’re going to ruffle some feathers with this one, but it needs to be done because too many founders are paying too much money to voluntarily handcuff themselves.

As the founder of a digital agency and a martech software company, I have a near-daily cadence of conversations with founders and customers that are concerned about their brand.

Of course, this concern is justified — everybody knows about the power of a recognizable brand. We experience it everyday when we pass a Starbucks, go food shopping, or walk through a pharmacy.

Before diving in on this, let me emphasize one important caveat — I’m not saying the idea of having a brand is BS, I’m saying the way that branding is done today (the process) is BS — particular for startups. When it comes to branding for startups, here’s a quick recap of what the typical process looks like:

1. Company decides they need to create a brand or refresh their existing brand.
2. Company hires Branding Experts (typically an agency or freelance designers) to execute the project.
3. Branding Experts quote a scope of services, which typically  ranges from $20,000 — $100,000+ (depending on scope of work, company size, and clout behind the experts).
4. Branding Experts begin execution through their own process/methodology, which usually includes interviewing Company for their preferences, some research, and showing Company several versions/iterations of where they’re headed to get their blessing.
5. Branding Experts unveil their output, which is almost always in the form of a deck (which is not meant to be edited). This product is effectively a specific and restrictive set of a rules that Company must always follow when making any creative assets.

Okay — looks pretty straight forward, and this might sometimes work well for larger brands with a lot of data and user feedback. But let’s take a deeper look.

When it comes to startups, we’re referring to companies that have small teams, little to no performance data, and haven’t yet hacked the product/market fit puzzle that could be a reliable lever for scale.

Let’s establish the real life-or-death importance of creative decisions: according to Nielsen, creative is by far the largest single contributor to sales success, amounting to a total of a 47% contribution — dwarfing other elements like targeting, reach, and brand recognition.

‍

If creative is so important, don’t we need to audit how we make creative decisions? Here’s the 3 major problems with our current system, followed by my solution to it;

Problem 1: Research — without looking at unique data from the company, the only information we can use to guide our decisions is ‘best practices’ (aka what works well for other companies). This generalized research, which revolves around overall industry trends, is what most folks look at for guidance in the absence of bespoke data, and to quantitatively justify their decisions.

This is known by academic researchers as ‘The Fallacy of Universal Best Practices’ (there’s a lot of research on this just a Google search away). Unfortunately, what might be good for larger companies can be disastrous for your company, and the highest-performing outcomes (the positive outliers) tend to come from practices that deviate substantially from best practices. Relying on best practices, or what works for others, is usually just as bad (if not, worse) than just guessing randomly.

Problem 2: Conflict of Interest — your designers care more about what you like rather than what your customers like. That’s because you’re the one directly paying their invoices, not your customers. I’ve never seen a creative production process where the creatives start by figuring out your customers’ preferences (at scale) — it always starts with finding out what the client prefers.

As much as we want to be empowered as ‘the client’ to have the final say and see our preferences come to life, we need to separate ourselves from our inherent biases. Based on the data we’ve seen at Marpipe, we found that what clients prefer is almost never the same (or even close) to what their customers prefer. To some clients, this is shocking and triggers a defense reaction, but to others, it empowers them to deliver creative that their customers really love. Which of these types of clients would you rather be?

Problem 3: Price — with a fraction of what you pay for a branding deck, you can create your own data on what your customers love, which will save you tens of thousands of dollars.

How can we do this? In short, we need to come up with as many brand hypotheses as we can (color palettes, messaging, graphics, images, etc), and test them all against our customers to see which variants they love and which ones they hate. To do this, you’ll need to run a multivariate test, which requires you to:

1. Create hundreds (to thousands) of ads that represent every permutation of those creative variables.
2. Run all of those ads to your target customers (with enough reach to hit statistical significance).
3. Component-analyze the data from all those ad results.

Doing this manually is pretty much impossible, and only accessible for large enterprises who can afford the resources to do it. However, with very recent developments in marketing technology, there are tools that can automate every step of this process for you — it cost less than $10,000 total (including ad spend) and only takes a few weeks from start to finish.

With a multivariate creative test, you can throw hundreds of hypothesis at your customers and gather deep intelligence on exactly what brand elements will create the most successful results for your company. With this data, you can even go as far as projecting what your Customer Acquisition Cost, Return on Ad Spend, and Click Through Rate will be under various branding scenarios.

My solution to today’s branding problem for startups?

Run a multivariate test and maintain a Modular Brand Book. With a multivariate test, you save tens of thousands of dollars while collecting custom data which tells you exactly which brand decisions will be most successful. Use this data to inform your initial branding guidelines.

Collect these guidelines in a brand book which is dynamic rather than static — expect for it to constantly change and be fluid as you test more and learn about what works and what doesn’t. Think of branding elements as “modular” — they are puzzle pieces that should be replaced over time as your company evolves. Multivariate testing should be done at least once per quarter, and the creative component data should drive the ongoing evolution of your Modular Brand Book.

This can free you from the ‘creative handcuffs’ of traditional brand guidelines, save you money while increasing your growth performance, and enable you to justify creative decisions with your own data rather than your own preference.