Posted on

Why your LLA’s are probably behaving badly, and why you should adapt.

Dimniko-Facebook-Lookalike-Audiences

LLA’s can be your best friend with media buying, by using Facebook’s algorithm to your advantage.

Recently we’ve seen that some accounts have seen a massive drop in performance with the LLA’s that have been working great and consistently in the past, it might be yours as well.

During this dreaded time of the pandemic, online customer behaviour has shifted for better and for worse depending on the industry you are in.

For example a study in Germany has shown that searches for luxury products have declined by 95% below normal levels in March 2020.

This is also including CTA’s, organic traffic and clicks to websites.

A recent survey cited from bigcommerce.com has found that 96% of Gen Z and Millenials are concerned about the pandemic and its effects on the economy.

“This concern is leading them to change their behavior more dramatically than other generations, which includes cutting back on spending, stocking up on items, and spending less on experiences.” – bigcommerce.com

With Gen X and Boomers, although still concerned about the pandemic and its effects on the economy.

The impact of their shopping behaviour has only changed by 24% of Boomers and 34% of Gen Z, compared to nearly half of Millennials.

Depending on your store’s main market, your Lookalike audiences can be your saving grace or your worst enemy.

This depends on how big the measures are that have been implemented in your target country and demographic.

Where there is a shift in overall online behaviour, your media buying behaviour should adjust and adapt to this as well.

That being said, this is how we made a shift in the strategy with accounts that have been showing symptoms.

Going back to the basics (again and again, rinse and repeat)

Many accounts have been relying heavy on LAL’s and using interest based audiences to test new prospecting customers.

Mostly we run the prospecting for a short term for testing and shift the budget back to LAL when we’ve seeded enough data into the pixel.

Here’s where the twist comes in, due to the change in behaviour, the strategy has shifted towards doing some good old fashioned research and focusing more on interests, demographics, age filters etc.

After this when we have enough conversions and the new data that comes with it, we populate the new LLA’s with the new conversion data, with a shorter conversion window period.

Prediction is key

As the market goes up and down, your strategy should ride the wave as well.

The reason why we see this weird decline and increase in the performance of these LLA’s during this period, is…

By the time you have enough data to have a valuable LLA, the market has already shifted again, due to a lockdown becoming more strict or being lifted, depending on the country and regulations around the industries, in it.

Which is most certainly the main culprit for this constant behaviour change.

The best way to track and prepare for this shift is to follow up on announcements from the countries you are targeting.

Onions have layers

From here, we broaden the interests based on data and some research and use a couple of new detailed targeting audiences and layer that onto your new LLA’s as the period extends.

This is where we start to extend the period of the LLA’s, and keep on seeding it to the adjusted online behaviour of your target audience.

We run new campaigns targeting new and previous interests and layer it with new and old lookalikes.

All and all, this is a volatile period for building profitable audiences that can give you that sweet sweet ROAS straight from the Top Funnel.

Adapting to your changing environment is key to building audiences that bring in the right amount of goodness to your company!

And this only works if you have the right agency partnered with you!

If you are spending over $500 a day and you want to scale your brand

Book a call below:

https://dimniko.com/msp-apply

And that’s a wrap!

I’m Quintin, Media Buyer at DimNiko

Posted on

LLAs Are Your Best Friends If You Want to Bring Amazing Results

DimNiko - Look Alike Audiences

We all know that different LLAs which Facebook can spit out

can work really great.

There are many sources from where you can create your LLAs

but for ecom we get best results from:

1- Past buyers
2- FB and IG engagers
3- Video viewers 95%
4- ATC or IC

We have several ways of how to test these.

Today I will show you one way

that brought very good results for one of our clients.

After the deep analyzes of whole account

we saw many good performing audiences at small segments of LLAs.

1,2,6,9% from PUR 30d custom audience.

Then we also saw some good performance from Worldwide audiences in ATC and VV95%.

So we did a testing strategy that went like this:

1-Test 1-10% LLAs in one campaign at $100 for US (source – 7d PUR)
2-Test 1-10% LLAs in one campaign at $100 for US (source – 30d PUR)
3-Test 1-10% LLAs in one campaign at $100 for US (source – 90d PUR)
4-Test 1-10% LLAs in one campaign at $100 for US (source – 365d PUR)
5-Test 1-10% LLAs in one campaign at $100 for US (source – EMAILS PUR)
6-Test 1-10% LLAs in one campaign at $100 for US (source – 180d FB eng)
7-Test 1-10% LLAs in one campaign at $100 for US (source – 180d FB vis)
8-Test 1-10% LLAs in one campaign at $100 for US (source – 180d FB eng_post)
9-Test 1-10% LLAs in one campaign at $100 for US (source – 180d IG eng)
10-Test 1-10% LLAs in one campaign at $100 for US (source – 180d IG vis)
11-Test 1-10% LLAs in one campaign at $100 for US (source – 180d IG eng_post)
12-Test 1-10% LLAs in one campaign at $100 for US (source – 30d FB eng)
13-Test 1-10% LLAs in one campaign at $100 for US (source – 30d FB vis)
14-Test 1-10% LLAs in one campaign at $100 for US (source – 30d FB eng_post)
15-Test 1-10% LLAs in one campaign at $100 for US (source – 30d IG eng)
16-Test 1-10% LLAs in one campaign at $100 for US (source – 30d IG vis)
17-Test 1-10% LLAs in one campaign at $100 for US (source – 30d IG eng_post)
18-Test 1-10% LLAs in one campaign at $100 for US (source – 75% VV)
19-Test 1-10% LLAs in one campaign at $100 for US (source – 95% VV)
20-Test 1-10% LLAs in one campaign at $100 for US (source – 180d ATC)
21-Test 1-10% LLAs in one campaign at $100 for US (source – 30d ATC)
22-Test 1-10% LLAs in one campaign at $100 for US (source – 180d IC)
23-Test 1-10% LLAs in one campaign at $100 for US (source – 30d IC)

We waited a little to get some results.

We didn’t launch all at once but we activate 3-5 test campaigns every 3 days.

The reason for that was so we didn’t put too much pressure on the current account budget.

Then round 2 of testing was even more testing based on the performance.

1- If we saw the best performance at PUR 180d, we also went to PUR 60,90 and 120.

2- At the same time we also logged in to our software we use for
better targeting, and created LLAs from 10-20%.

3- Then we also grouped best performing LLAs into a big mix
where we combined all 1%, all 2%, all 3%, etc. from different sources.

We waited a little after that, optimized and launched a few more when we saw light.

There is alway some quality lighting at the end of our ‘tunnel’,

so at the end we knew we have to try also Worldwide lookalikes.

We created all above LLAs from round 1 and used different regions

and countries such as the UK,EUR,AU,UK,etc.

It started to become really messy in our audiences account

so we had to put everything in excel to organize our tests and have a better

view on what is working and what not.

We made a detailed plan for the next few weeks

and slowly pushed out different tests.

In round 3 interesting was, that we found some good combinations

from worldwide LLAs but targeting just US or a specific country.

This is where everybody said: “Ok, that’s enough!

We have enough small winners that we can make bigger winners”,

and we did one more thing.

We mixed all our best performers from round 1 and 2 into a very big campaigns,

intersected with very broad niche interests (beauty, online shopping, yoga)

and turned on detailed targeting expansion.

We got some amazing campaigns that we were able

to push to high budgets immediately.

And at this phase all the previous testing paid out.

Many of you will never want to go this way because it gets pretty messy,

but with right strategy, you can bring to your clients

(or to your business) amazing results.

And we always want to bring the best!

That’s why our clients love us. 🙂

So if you’re spending over $500 a day

And want to scale your brand

Book a call with us below:

https://dimniko.com/msp-apply

Have a great day,
Matej