Can YouTube Ads work on a small budget—say, $50 a day? And if so, how should you approach testing it?

Our take: absolutely, 100% yes. While the measurement side isn’t highly sophisticated, it’s perfectly appropriate for clients at this level of spend and revenue. It’s scrappy, but effective.

We were all excited when Haus released its latest study of just how incremental YouTube is for their partners. If you haven’t checked it out, click here. I highly recommended reading it.

However, as I was reading the blog, I began thinking more about some of our partners with smaller spend levels. The ones who can test with spend 2.5k - 10k per month, how should they go about structuring a YouTube test?

How can you test YouTube with $50 per day?

Below you’ll find the context of our latest test including how we structured the test & how we’re measuring results.

Context:

Recently, we pitched a client on running YouTube Ads to help expand our reach and bring in new customers. We were spending about 10k - 12k a month on Google Ads, all focused on Google Shopping & Search. Everything was performing at goal with most spend dedicated to non-brand and new customers. The problem was that this particular category of product(s) is niche and we were maxing out our spend.

As opposed to keeping things status quo, we pitched this client on opening up a small YouTube test to prove out the channel as viable driver of new customers. The client approved $50 per day.

So the big question is can you make $50 per day work on YouTube?

The answer is yes and here is how you do it.

Campaign Structure:

When testing YouTube at Banfana, we like to run matched market / GEO lift tests. These tests have their detractors but again, at this spend level you don't have a lot of options. Specifically, this is where we pick a variety of DMAs to run YouTube in and measure them against control markets.

Usually, depending on the budget, we pick a host of DMAs as our test markets, match them with similar DMAs as our control markets & run a 60 day test with a 30 day cooling off period. Then we measure results. This is nothing novel, companies like Haus can do the same thing. However with smaller budgets, it can be more difficult to measure statistically significant data.

In this case with our budget of $50 per day, we could only pick one DMA to test in. I want to acknowledge that with smaller budgets and one DMA, you do run the risk of getting misleading data. But with this client, we had 3 things that made us confident in YouTube:

• Creative variety. UGC, Founder Videos, Product Focused. Tons of content to test that was already performing on Meta & TikTok.
• Easy measurement. 2 places where the consumer can purchase; Shopify & Amazon. Most transactions happening on Shopify.
• Specific Audience & Product Market Fit. Audience targeting condensed into clear segments with specific Demographics (gender & age range).

With confidence in our audiences, creative, & measurement, we were happy to proceed with a matched market lift test.

Measurement & Results:

Our process for measuring the results in this test follows these steps.

First, go into Google Ads > When & Where Your Ads Are Shown > Matched Locations > Cities. That way I can see exactly where the increase in traffic is coming from.

Matched Location report from our YouTube Test ^

‍

Then I like to go to Shopify > Customers By Location report. I filter for shipping city and match those cities with the report from Google Ads. I then overlay the Shopify > Sessions By Location report so I can see both traffic uplift and sales uplift (fingers crossed for the latter 🤞).

‍

The entire test is 60 days of spend followed by 30 days of the campaign being paused. This gives us enough time to evaluate the uplift. Sometimes, results come in much quicker. I personally really start to dig into results after the first 30 days. In this case, I waited a bit longer because of the spend levels were smaller.

What we saw with $50 per day, targeting one DMA was quite incredible 🙌

Here are the stats in our targeted DMA:

• Net Sales were up 395% vs previous period AND 368% vs Shopify
• Sessions were up 85% vs previous period AND 96% vs Shopify
• iROAS was 2.91x driven almost entirely by new customers

No results sharing without caveats.

1. We’re measuring a non-sale period vs non-sale period.
2. We’re targeting one DMA so the risk of being misled either positively or negatively by the data is higher.
3. You have to match your spend level & potential uplift to your current DMA session & sales level. In this case, the data is pretty clear but in others, you might see a significant shift due to under investment.

Next steps:

We’re doubling our budget and expanding our DMA test to one other lucky location. Our creative pipeline is expanding as well and we’re adding more videos to the mix.

Can you spend $50 a day on YouTube and find success? The answer is yes! You can absolutely make smaller budgets work. However, I would reread the 3 things we needed here and the caveats for context.

Hope you enjoyed reading and happy scaling 🤘