10 Secrets For Successful Facebook & Instagram Ad Campaigns

Leaf Grow - Blog - 10 Secrets For Successful Facebook and Instagram Ad Campaigns.png

These types of blogs usually point out the glaringly obvious and don’t give much in the way of insight into campaign execution mechanics. I’ve tried to avoid the former and do more of the latter, but some obvious things are worth a little reminder.


#1 Align your decisions with your objective

If there is an overarching message to this blog, it’s this. Every good decision you make will stem from having a structured ad campaign plan that’s designed to help you achieve your objective(s). It sounds blatantly obvious I know, but it’s mind blowing how many people/businesses don’t get this advertising fundamental right. Example:

If your objective is to drive traffic to a destination, make sure you’re optimising for this. Don’t run a video promotion campaign and just stick a link in the ad copy and hope people will click on it. You’ll get very few clicks, because you’re optimising for views. Inversely, don’t run a link clicks campaign when what your really want are views and engagement on your video.

This doesn’t just apply to your campaign type, it’s a seam that runs through creative production, audience targeting decisions and the way you measure success.


#2 Know the funnel, know the journey, know what you’re optimising for

Having an intimate knowledge of your sales funnel and/or the customer journey is crucial. Break it up into manageable stages and tackle them one at a time, from top to bottom. Implement the basic foundations of the whole customer journey so that there’s no major blockers on the road - a shit checkout experience where you struggle to even find the cart or purchase button is one such example - and then start to sculpt the different phases that make up the journey.

It’s no use setting up a conversions campaign and optimising for purchases (bottom of the funnel) when people are struggling to log in to your platform, or are waiting too long for your landing page to load. Get people into the top of your funnel and get your data-feedback-loop whirring away and then start optimising: conversion rate from from ad to landing page views; conversion rate from landing page views to signing up for a trial etc. Once you’re seeing the desired level of gains in one phase of the journey, move on to the next part and tweak that until you see gains there. This is a constant process - things can always be improved.

Lots of tiny tweaks equal one big change.

#3 Use the right content, in the right place, with data-led design

Usually these types of lists will include something lame like “make great content”, but that’s such a nebulous statement and means so many different things to different people. In our experience running thousands of FB & IG ad campaigns, it isn’t always the greatest (most beautiful or professional looking content) that performs best and now people’s “ad radars” are so well tuned that they can spot an ad in a nano second and subconsciously filter it as they scroll on by.

Sadly (or not) great content is content that works to help you achieve your objective, not how conceptually clever, meaningful or beautifully designed it is. Function can be the form. That’s ok.

Test all placements to start with and once you see which placement is consistently delivering results then optimise your ad designs for this placement.

Don’t assume that one particular channel or placement is going to work best for your campaign until you have enough data to know this for a fact.

#4 Let data validate your instincts - too narrow, too early

Assumptions are essential, they’re the offspring of ideas and the starting point for all your marketing endeavours. Unless you have a glut of irrefutable data (over a decent time period) that tells you exactly who your customer/fan is and what channel they principally live in, then don’t let your assumptions narrow your focus too early.

If you’ve built a picture of your key audience persona and you’re confident you know who your target market is, great, use that as your ‘anchor assumption’ and then spread the net a little wider around that.

If you think your core audience is 25-34 yr old males who reside in Western Europe and live on Instagram, then start by targeting 18-44 years, both genders, include Southern Europe in your territory group and target Facebook and Messenger too. This way you’re validating your assumptions.

If you only target the exact people who you think make up your target market then you’ll never know if there’s an opportunity just beyond those boundaries.

#5 Have a territory targeting strategy

Sorry to keep banging on about it but deciding which territories to target for your campaign should be led by your objective.

If, as a music marketer, the objective is to drive fans to streaming platforms, then target regions that have decent premium streaming penetration levels and a penchant for the type of music you’re promoting. Be mindful of the territories in which your other marketing endeavours are active, such as traditional PR, radio and TV etc. As As a retailer, it might not make sense to ship to certain territories due to transport costs and/or and import tariffs so there’s no point including these in your targeting. Other regions have profoundly different cultural nuances and practices that you could never learn from simple market research. If you’re unsure then err on the side of caution and leave these alone until you’re confident you have reliable local knowledge.

There are major pitfalls and opportunities when choosing your target territories and these can fundamentally affect the success of your campaign. Especially if you’re not paying attention to the performance data.

Certain territories are more expensive to advertise in than others, which can easily mean having your budget emptied into a region which isn’t a priority.

#6 Use a daily budget and apply it with care

Daily budgets allow for more control and flexibility than lifetime budgets. If you're setting up a trial campaign for a client, where your goal is to reach a certain cost-per-result that will unlock an ongoing relationship, then it’s vital that you use a daily budget. If you use a lifetime budget and you reach the desired cost-per-result through 6 weeks of optimisation and the campaign comes to an end, you’re going to have to start all over again and there’s no guarantee that a carbon copy campaign will work just as well. There’s too many variables at play and campaign performance is a delicate thing. It’s going to take you a while to get down to the optimised results you had before, and by this point the client’s patience will have vanished, along with any future income.

You don’t need to wax large sums of money on your campaign straight away to start getting good trend data. In fact this would be the worst thing you could do. The minimum daily spend in Facebook is £2. This will give you 2 ad sets. It’s one pound per ad set, per day, thereafter. You can test a large range of content against a large range of audiences with a very modest budget.

Start small, and as you begin to uncover a results trend then increase your budget incrementally to test the stability of the trend. Fifty conversion events is the ideal number to reach for stability, although anything over 20 should be a safe enough bet.

Only hockey-stick your budget when you’re confident the trend you’re seeing is stable enough to handle more dramatic increases in budget.

#7 Test one thing at once

There are so many variables to Facebook and Instagram campaigns it can be overwhelming. Again, there is a sweet spot between ‘too narrow’ and ‘too wide’ in terms of targeting and this depends on what you're promoting and what your objective is.

Starting your campaign with proportionately wide targeting parameters feeds Facebook with a healthy dose of data to analyse and act upon. In the early stages of your campaign don’t worry too much about ad channels, placements and devices, you can always tackle these once you know what kind of image, messaging and audiences are working. This ties back to your funnel, knowing what your optimising for and using content that works, based on data.

As a simple guide:

  • Start out by testing messaging variants with a handful of different image types (make sure the image types are distinctly different). This will enable you to identify a solid trend in messaging that works across different image types, therefore highlighting that it's the message that’s driving results.

  • Hopefully you’ll also be able to see a trend on image type here too. Take the best performing message(s) and create variations of the best performing image type to see if you can improve upon the results by doubling down on the concepts that are working.

  • Now you have your best-performing message and image type working away, maybe start testing images against videos, then move onto placements and so on.

It’s all about identifying single variables that are working and then iterating on those to give you percentage gains.

#8 Measure what’s important

We have a habit of placing importance on the things we can measure easily, to make things look good, instead of figuring out what’s important and measuring that. As a slightly macabre example, the American government measured and reported their success in Vietnam based on enemy body count. It was the only constant metric they could measure. They kept throwing money and human resource at it for years and the body count kept on rising. We all know how that worked out.

I hear a lot, especially from the world of agencies, about reach and impressions, but these mean very little in isolation. They’re just easy to measure and look good because the numbers are high. The same can be said for traffic campaigns - you might be driving a lot of traffic to your website and the cost per click might be super low, but what’s the bounce rate and the dwell time? Where are these clicks coming from? What’s the conversion rate to sales or signups?

Concentrate your focus on high-value-actions that help you achieve your objective.

#9 Interest audience development

Finely tuned interest-based audiences often out-perform lookalike audiences and what’s more, you know exactly what’s going on inside them. Lookalikes are useful, but murkier, so don’t give you as clear a window into what your core audience looks like.

You can test hundreds of interest audiences, see how they perform and start to identify very high performing ‘super interests’ for your brand/product.

Read our blog on Super Interest Building, Testing and Analysing to see how to do this.


#10 Experiment. Measure. Refine. Repeat.

Every campaign you run is an experiment and even your failed experiments are still incredibly useful; they tell you what not to do.

When you hear successful entrepreneurs get asked what made them successful they usually say something vague like “trusting my gut”, “being lucky” or “having great people around me”. If you ask the same people why one of their ventures failed, they’ll be able to tell you the exact actions, times and dates that led to failure. They know exactly what not to do, but they couldn’t give you a blueprint for success.

There is no silver bullet to Facebook and Instagram ad campaigns, but understanding and using the right techniques and following best-practice will help you confidently steer through the waters. These techniques and best-practices are meant to be experimented with, refined and improved to make them work for you.

Finally, if you’re not measuring the outputs and outcomes of your marketing actions, you won’t know what’s working and any improvements you make will be by sheer luck, which isn’t repeatable or scalable and as we all know, luck doesn’t last forever.