Welcome to Part 2 of Fighting the Facebook algorithm. In Part 1 I gave you a few tips on how to get your messaging out, how to balance that sales pitch with effective content.
Now heres the thing, I’ve read many Social Media books, attended many presentations and read loads of blogs. All of them talk about quality content. The things is they never really explain how you define what that is.
Every audience, channel, time of day, day of the week defines what is hard working quality for you. there is no set formula. Although there is one common type of post that just seems to kick ass – something I’ll mention further down – just to make sure you are reading this.
The first thing you need to do is test. Test more and keep testing. Even when you get to a point that you have data to take learnings from, you keep testing. Organic is just that, organic. It is constantly changing, you have to make sure you are on top of whats working.
So where do we start the Facebook Algorithm battle? – break your content into groups
You need to break your content up into groups. Now this is your call what your groups are as it will differ across pages. Here’s an example :
Lets say you are a restaurant. Every week you post images of your dishes, occasionally you post a staff photo, now and again you post an Instagram one of your staff took, you also post a special of the day and occasionally you run a competition.
So you create content groups like this :
- Staff
- Food
- Competition
Still with me? You try to find some common ground within the posts you have up. It could be more than three. (don’t over do it though) . You should have an idea of the type of stuff you post, just give them a common term.
Now you need to get to the data, I use Sprout Social (affiliate link) for this, however you could get away with using Facebook insights.
Now I want to filter my posts by the most reach. Click the header title ‘Reach’ – Facebook insights or Sprout social should list your posts in relation to that.
Now take a screen-grab of your report, take as many as you need in order to get your data.
Next, open powerpoint and add each one of those screen grabs to a slide.
Now the fun part, you get to colour in. Remember those groups you defined earlier. At them as colour codes to your slides. Like this :
Now start colour coding each and every post according to your split. Like this :
See what we are doing here? Slowly but surely we should start to see what types of post are spending the most time at the top for reach. In this case we can see competitions are the highest performing. Although I put that in there for a reason. Competitions will always come out on top. There are a lot of people looking for that free thing. So what I do in this case, remove competition posts from the equation and work with whats left. The reason being, you know in most cases, a competition will get the reach for you. Only problem is, a majority of the users that will interact with it or follow you because of it are just there for the competition. You won’t hear much from them once its ended.
Make sure you have discounted any paid activity here by the way, this is all about organic reach – If a post was supported by some paid activity, it doesn’t count)
Now to figure out what media types are working
So now you have a few slides of content groups all colour coded, you’ve taken any competitions and paid activity out of the question. Take a step back and see the most frequency and highest appearances of high reach post types. In this example posts talking about food are the clear winner. With Staff mentions coming in second. (Obviously you would have a few more groups but you get the idea) I would aim for no less than 5 but no more than 10 groups.
Fantastic, so far we know that the highest reaching posts on our Facebook page are Food and Staff. But wait theres still more to do, we may know the content types, but what about the media types? That is, were they videos? Galleries? Text? Images? what worked? The reason you should do this is because food posts may beat Staff posts, but only the video versions. Or Staff posts may beet food if they were a video post.
You need to now go into each content type and see what types of media are working. Do it like the above, but this time break the high performers in to new groups :
- Video
- Text
- Gallery
- Image
- Text
- Audio
Once you have done that you will start to see little rules for your content. For example :
- Food needs to be in Galleries and don’t work on their own
- Staff posts need to be accompanied by an image
Still with me?
Whats the best day to post?
The next thing is something thats more a guide than a rule, but it does help. You need to know what day works.
First, take a break. You’ve done a lot of reading so far. This Facebook Algorithm process is far from over
Right, welcome back.
Day of the week is handy as it helps you plan ahead and get the better results. This is a bit of a boring task, as you need to look at that original reach report again and start noting dates. Mark them on a calendar or note book, just get to a point where you can see a sort of trend. Are Wednesdays a busy day? Is there hardly any reach on Sundays?
You may want to take note of the times posted, but I have another opinion on that later.
In the above example I can see the top performers were Monday, Tuesday, Wednesday and Sunday. So now you need to make sure, using your already defined rules we established earlier to test the high performing posts on high performing days. These ones :
- Food needs to be in Galleries and don’t work on their own
- Staff posts need to be accompanied by an image
You need to plan more posts with these rules on some of those high performing days so you can get more learnings. You want to get enough data so you now start creating rules :
- Food needs to be in Galleries and posted on tuesdays
- Staff posts need to be accompanied by an image on Sundays
How cool is that, you know have a list of simple rules to use when planning your content. This is going to make filling in your Content Plan so much easier.
Earlier I mentioned about noting down the time of post. I don’t do this anymore. In Facebook’s latest algorithm word is they removed the hour factor. Originally if you posted at 14:00 that post may be displayed to your most engaged user, but they might not go onto Facebook till about 16:00 – in those two hours, other posts would have pushed yours further down. What Facebook does now is display posts regardless of time posted to users that seem to want more of your content. You could post it at 13:00 and that user checks Facebook at 17:00 – if they are highly engaged with you that post will appear higher up than other posts in their feed. So don’t worry too much about time of day. General rule, working hours.
Finally I said earlier about the format on Facebook that works amazing every-time. Galleries, I shit you not. every single client of mine has great performance off the back of galleries. Make sure you include them – but don’t over do it. I keep it to one a week.
There you go, part 2 done. Hope it helps. Drop me any questions in the comments below. This system has helped me ensure my clients Facebook pages are working hard and every post is well thought through and serves a purpose. Not just posting for the sake of posting. It takes time, but once its all in place, it just works. Just create rules based on the data available to you.
If you don’t have few months worth of data or different types. This is where you test, and keep testing. Post a video in the morning on a Monday, then post it again on a friday evening. (It’s okay to repeat the occasional post – remember, the majority of users don’t come back to your page, they only see whats in the news feed) I find it takes about two months of data to start making some informed decisions. Just post variants and start collecting data so you can start applying the above
In part three I’ll go over using paid advertising to help you keep that balance and counter the effects the Facebook Algorithm has on your page. It’s not about blitzing money, its about strategically choosing times to attack with money.
See you then!
