Tips and tricks for paid social & acquisition

An interview with George MacGill

Hey everyone,

Below is my interview with George MacGill. I used to work with George at Social Chain and he was always one of the sharpest thinkers there on multiple topics.

He is super knowledgeable on performance marketing and paid social so I wanted to pick his brains a bit more here to help you guys whether you are starting a brand, working at an agency, a big company etc. Hopefully some of the tips can feed into your own ad plans.

You can catch more of George on Twitter (29.7k followers) where he talks more on mental models and e-com and on his own newsletter - Return of The Mack!

His LinkedIn details are also here.

Let’s get to it…

What are the biggest platform changes in the past 12 months that have affected paid social? And how did COVID shift things for you?

One of paid social’s biggest drivers is e-commerce.

When you see 10 year changes happen in 8 weeks, it has knock on effects on the rest of the ecosystem.

The periods of March - July were by far the most lucrative we’ve ever seen from an acquisition perspective.

With a changing climate everywhere due to external events, it’s difficult to draw causation and correlation with changes on the platform. 

Paid Social / Ecommerce in general is one of the few fortunate industries (alongside video conferencing, hand sanitising and face masks) that has weirdly benefited from an awful crisis. 

Are there platforms outside of Facebook’s ecosystem you like at the moment for paid acquisition?

YES!

Snap has been a personal favourite for a while. However, it skews to a younger US audience, so only makes sense with the right product. 

They did what Twitter failed to do and built an ad product that actually competes with Facebook.

From a UI + customer service perspective, it’s ad product is miles better than Facebook’s IMO. 

If you have any issues with the ad account, you can chat to a friendly customer service rep immediately. 

The UI is tremendous and clean - especially for beginner media buyers.

If you contrast this with Facebook’s, it wins night and day in this regard. 

However, Facebook (with Instagram) still holds the top spot due to its customer base being older with a larger amount of disposable income.

I’m in the process of experimenting with TikTok ads, where there has been some mixed results. If the platform continues to grow and doesn’t stagnate under Oracle, I can see the ad product being up there with FB and Snap in the next 12-24 months.

YouTube for higher margin products is a platform I expect more brands to invest heavily into as they look to decentralise from Facebook. 

What do you think are the biggest arbitrage opportunities in paid social at the moment? Where is attention cheap? 

Good question.

Assuming you have solid ad buying practices and do the fundamentals, then creatives are by the far biggest exponent on paid social performance.

Quantity and quality is key here. The more you have and the higher quality they are - the better your performance will be.

The key here is to have the greatest variety in the creatives. Having 10 similar creatives isn’t useful at all. Having 10 unique creatives is paramount. 80% may fail - but you will find 1 gem, and often that gem isn’t the one you would’ve assumed will perform. This is why testing a variety is so important.

In terms of other opportunities, I’d say the following: 

1. Using influencer accounts at top of the funnel

- Get advertiser access permission from the influencer
- Run the ads from their account

^^ A few brands have started to do this in the last year, it’s unique and stands out. We are used to seeing sponsored posts from brands, so when you see it from an influencer - it is a pattern interrupt.

Having the ability to target influencer audiences and influencer lookalike audiences reduces the uncertainty of reach via Instagram’s algorithm.

Even more notable, using influencer and influencer lookalike audience in targeting makes your influencer’s follower count irrelevant.

Custom audience targeting provides you with the ability to reach more of the people you know are interested in what you’re selling..

2. Digital Products

This isn’t anything new. Make Money online gurus have been doing it for years.

However, in the same way the internet and bitcoin started off with a shady reputation (the first thing sold on the internet was weed), I think this industry will really transition in the next 10 years. 

When Netflix costs £71.99 per year to stream videos vs University of Salford’s £9,250, it means there’s going to be huge demand for alternative education.

The advantage that digital products have from a paid social model is a 100% margin, which means that if they only achieve a 1.5x ROAS - it is an incredibly healthy ad campaign.

In contrast, most physical products don’t break even on the front end until 2.5x> ROAS (product depending).

Are there any tips you would have for people running campaigns during Black Friday/Cyber Monday when CPMs are elevated and attention is crowded? E.g. do 24 hour product drops vs. heavy discounts? 

The biggest tip I would have is to start using ads to build out much cheaper channels on these days. So I would build out your email/SMS list as big as possible for that day - so you can use cheaper acquisition means than just using ads as a way of relying on sales.

ROB: Most brands use BFCM as a way to increase cash reserves and get a loss on taxes before the end of the year which is why most brands run such aggressive sales. When you’re giving product away it’s not particularly hard to push big revenue numbers, when you’re losing money (not making a net profit).

Other stuff I’ve seen work personally includes:

• Re-releases - take products that sold out in the past and do a one time only re-release giving customers one last shot at buying that product they missed out on.

• New product drops - Drop a new product or line on Black Friday. My Protein example below.

• BFCM-only merch - Black Friday only merch can be provided as a gift if a certain cart value is hit. For example, if you’re in apparel, make a high-end hat that is gifted for customers that spend over $200. While the hat only costs ~$7 or so to manufacture, the customer perceived value might be $50+. The customer would much rather have the hat and you’d rather have the $200+ than discount. It’s win win.

  

Using the above methods, one small company I know, achieved $200k+ in sales over BFCM in 2019 but maintained their 45-50% margins.

Scarcity drives higher sales vs. discounting like 99% of people do, you can discount at a completely different time of year to stand out if you want to e.g. early October.

What are some of the most unique customer acquisition & lead gen campaigns you have seen or run yourself in the past 6 months? 

GEORGE: There’s ones I can’t give away publicly! Ha.

However, one of my favourites that I ran for myself are hyper specific campaigns based on personalisation.
For example, a nationwide recruitment campaign targeting everyone in the UK. 

Flip this on it’s head and make small adjustments to each ad for graduates of each specific university.

So rather than UK graduates, it’s University of Leeds graduates. That one small change of personalisation saw a massive uplift in performance.

ROB: I.e. The Gary Vaynerchuk approach. Getting really specific with content, he has advocated in the past doing as many pieces of content (as makes sense) to one ad set. So for example, if you are selling apparel for NFL fans, this is probably far too broad, so you should niche down to selling to Jets’ fans and ads speaking in their language (humour, tone, historical knowledge, design preferences).

It’s the same principle of finding a smaller niche first (I mentioned anime girls smoking a few weeks back and the sub-reddit for testing that) just applied to social ads. I first learnt this lesson back from Tim Ferriss’ ‘4 Hour Work Week’ where he talks about the extreme example of dwarf entertainment rentals but have seen it crop up time and time again in different ways in e-commerce and paid social.

GEORGE: Another cool hack that Pet Lab are doing is building a community around their product with Facebook Groups.

There’s over 60,000+ customers in their group talking about their product - which is insane. 

ROB: Peloton doing the same thing with 345,000 FB Group members, also a sub-reddit with 91,000+ members. When fanatical communities develop around your brand you’re off to the moon!

What would be your biggest piece of advice for someone starting a brand today for initial customer acquisition from paid? Beyond just CAC < LTV and the basics from Facebook Blueprint

GEORGE: Hmm…

It would be so contextual depending on the brand. So I will go with some of the broader answers:

1. Pay an expert to get you up and running with the basics.

The money you will earn back from it will far outweigh the money you will lose from it.

2. Get 5-10 different creatives made

This could be user generated content, animations, product unboxing etc etc

I’d use Facebook ad library and combine it with your own brand guidelines to create something unique. 

3. Track the following metrics:

A. Cost per unique landing page view
B. Conversion rate
C. Average order value

After your first round of tests, you will realise that one of these is the biggest area of growth. Adjust and adapt based off which one it is.

For direct response campaigns, if you had to give 5 pieces of advice for better conversions, what would you say? 

1. Invest in your landing page
2. Invest in more creatives
3. Who can pay the highest CPA will win (So make sure your AOV (average order value) / LTV is maxed out)
4. Be scientific with your testing (Don’t have too many variables running at once)
5. Have other channels other than direct response (It will help fuel your direct response channels better)

Getting more specific, can you give us an example of a couple of top of funnel setups? E.g. CBO -> 1% lookalikes view content or purchase, 3-4 ads or?

To get more nuanced, it would be budget dependent. But if we assume they are spending £1,000 per day, I would suggest each ad having an 1x AOV of spend per day to know how well it is performing.

So if the AOV is £100, you’d have ~10 ads live that day. If the AOV is £10, you’d have ~100 ads live that day.

You would then work your campaigns and targeting based off the number of ads you anticipate having live. 

So if we have 10 ads live, 3-4 targeting ad options would make sense here.

I’d suggest testing everything as no ad account is the same but if I had to chose 3-4 ad sets:

A. 1% lookalike of previous purchasers.
B. Macro interest - millions of people.
C. Micro interest - 75,000 - 1 million people.

This gives enough optionality for you to get more data / insight. 

Can you also provide an example of a full funnel setup for a client or two that is working very well at the moment (high ROAS/low CAC)?

35% of the budget towards testing.
35% of the budget towards scaling.
10% of the budget towards retargeting warm customers (engaged with content + seen ad)
10% of the budget towards website retargeting
10% of the budget towards past customers that are looking likely to churn.