Search & Gather, a small digital marketing agency in Toronto, led yesterday’s Growth Hacking Toronto session about running paid search and social campaigns. These are my notes from their presentation.
Choosing the right platform means understanding the product (or service) you’re promoting and the audience you’re promoting it to.
You may have some assumptions about the audience — who they are, what they want, where they hang out, and what platforms would be the most effective — but those are only assumptions.
What does the data say? We’ll come back to this in a moment.
The type of CTA you use — the call to action, e.g. “sign up” or “buy now” or “learn more” — will also affect your platform choice. Some CTAs are better suited to certain platforms.
Ditto for when the conversions — the action of clicking on the CTA — actually happen.
Targeting an evening behaviour (e.g. finding a place to grab a drink with friends) may be better served on a different platform than targeting a daytime behaviour (e.g. scheduling a business meeting).
You’ll have assumptions about the target audience. You’ll have assumptions about the CTAs. You’ll have assumptions about timing.
Test your assumptions. Run trial campaigns and see what the data tells you.
Start with a small budget and iterate your tests across 3-4 platforms. As you get comfortable and grow your confidence, gradually increase your spend.
If you have a tight budget, use fewer platforms so you’re pulling in statistically significant data. (Search & Gather recommended a minimum of $1k per campaign.)
So, which platforms should you use?
LinkedIn is good for targeting industries and professions. Unfortunately it has a bad algorithm.
Bing is like “AdWords little brother”. It has lower CPC and higher ROAS. (I think they mentioned 2x to 3x return.)
WeChat is a great way to reach the Asian market.
Custom Audiences, Lookalike Audiences, and Remarketing are all underutilized on Facebook.
An additional point on Lookalike Audiences: Base your audiences on different levels of your marketing + sales funnel. Start with 1% lookalike and “work out from there”. To reduce the budget, layer on geographic and demographic filters.
Call Campaigns (advertising a phone number – mobile only), negative keywords, and key positioning with better creative (e.g. targeting 2nd or 3rd placement but having a better offer) are underutilized on Google.
Sticking with Google: Dynamic search ads can get bring in traffic from longtail keywords, and perform well with a lower CPA. Google Display Network offers advanced targeting. Start with responsive ads.
Twitter is high maintenance. (Yes, that’s the only note I have about it – conversation was mostly focused on FB and Google.)
Audience fatigue is real
Audience fatigue/frequency fatigue happens when the same audience is exposed to the same advertisements and messaging over and over and over.
To get around it, rotate your audiences. Segment your “big audience” into smaller groups. Alternate your spend on-and-off to target the different segments at different times.
You can also test new audiences and add them to your rotation.
Bottom line? Don’t waste your budget on a tired audience.
Let’s talk ad creative
There’s still a ripe opportunity for good, interesting creative in online advertising. Channel your inner Don Draper.
If you’re not sure which direction to go in with your creative, start simple. “Keep it vanilla.”
From there, experiment and iterate with your creative, the same way you iterate with your targeting.
Run multivariate tests. Try different messaging and images. “It’s not set it and forget it.”
Continuity & relevance between ads and landing pages
So a prospective customer clicks your ad. What now? What do they see on the landing page?
Should the creative be the exact same as the ad? Should it be different? Does it matter?
“It’s situational.” Sometimes the creative doesn’t need to match 100%. But the tone and messaging should be consistent.
What about video?
“Video is hard and expensive.”
Traditional brands are doing it well because they have the cash to spend on production.
Small businesses and startups don’t have that kind of money. So they go static. It’s easier to test and iterate and “fail fast” on images and text.
…and that’s it! Thanks to Search & Gather for sharing their knowledge with the group and putting up with the inconveniently-timed fire alarm tests.