Analyze to improve your next post
Analytics is not for staring at numbers and feeling relieved or discouraged. The purpose of analytics is to make the next post you publish one step better. You look at the numbers, find clues about what worked and what did not, and apply them to your next post. That cycle is analytics.
So you can look at only a few numbers. If you juggle this with other work, limiting yourself to what you can review in 30 minutes once a month is more likely to last. Analytics that does not last will not change your next post, however carefully it is done.
Decide your goal first
Which numbers to look at depends on what you want social media to achieve. A shop that wants more visits and a company that wants more job applicants care about different numbers. Here is how goals and numbers relate.
| Goal | Main numbers | Numbers to check alongside |
|---|---|---|
| Get known | Impressions, reach | Change in followers |
| Spark interest | Engagement (likes, replies, shares, saves) | Engagement as a share of impressions |
| Prompt action | Link clicks, inquiries, visits | Inquiries as a share of clicks |
| Keep the connection | Change in followers, number of replies | People who engage repeatedly |
Once the goal is set, the numbers to watch narrow to two or three. How to set goals and choose the numbers to aim for is explained in how to set goals and KPIs.
What the main numbers mean
Impressions and reach
Impressions are the number of times a post was shown on screen. If it is shown twice to the same person, it counts as 2. Reach is the number of people who saw the post, and the same person counts as 1. Which of the two a network reports differs by network.
Engagement
Engagement is the number of actions people took on a post, such as likes, replies, shares, saves and clicks. Which actions are included differs by network, so do not compare numbers from different networks directly. Engagement as a share of impressions is covered in detail in how to calculate and read engagement rate.
Change in followers
Look at follower counts as the difference from month to month, not the daily ups and downs. Comparing what you posted in a month when followers jumped and a month when they dropped shows which posts fit.
Clicks and what happens next
Link clicks are the entrance to action outside the social network. How many inquiries or purchases came from those clicks cannot be seen from the network's numbers alone. How to tag links and count on your website side is explained in measuring results with UTM and short URLs.
The order to look in
Going from large to small keeps you from getting lost: the whole account, then individual posts, then time of day and format.
First, compare the whole account's numbers with the previous month. Just see whether impressions, engagement and followers went up or down. Next, sort individual posts by engagement. Put the top 3 and bottom 3 side by side and write down what was different, and you will see what to change in your next posts. Last, check whether engagement differs by day of week, time of day, and with or without images or video. Thirty minutes is enough for all of this.
Cautions when comparing
When you compare posts, match the conditions. If you compare a post from a few minutes ago with one from a month ago, the older one looks bigger. Compare posts that have been out for at least a few days. Also look at the number per post, not only totals, so that one big hit does not mislead you.
What to do after looking
After looking at the numbers, decide on just one thing to change in next month's posts. For example, "The top posts were photos of hands at work, so we'll add one such photo a week next month," or "Tuesday evenings got more reactions, so we'll move one weekly post to Tuesday evening." Narrowing to one lets you check next month whether the change worked. If you change several things at once, you cannot tell what worked.
Keep a short record of what you saw and what you will change. If you share results inside the company or with clients, using the same report format every month makes comparison with the previous month easier. How to make one is covered in how to build a monthly report.
Common misreadings
Misread numbers can lead you to the opposite of the right action. Here are three that often happen with small business accounts.
Treating one big hit as a "winning formula"
When one post does 10 times better than usual, it is tempting to keep posting the same type. But if it spread because a well-known account happened to quote it, or because it coincided with a topic of the day, the same format will not repeat it. Judge by whether several similar posts, say 3 to 5, do well on average, not by one result.
Taking growth in totals as proof that your work is paying off
Even if total impressions rose, if you simply posted more, the number per post may actually have dropped. In months when your post count changed, look at the per-post number together with the total. For businesses where people's behavior changes with the season (year-end gifts, spring moves and so on), comparing with the same month last year helps separate the effect of the season from the effect of your efforts.
Looking only at what can be measured
The numbers on a network's analytics screen cover only what happened inside that network. Results outside it, such as store visits or phone inquiries, do not show up. Don't treat results that don't show on screen as if they never happened. Set up at least one way to count outside social media, such as a shop asking "What did you see that brought you in?" or a "How did you hear about us?" field on your inquiry form.
What you can do in SNS TOWN
Analytics in SNS TOWN puts the numbers for your connected X, Instagram, Facebook Page and YouTube accounts on one screen. At the top, there are line charts with followers on the left and engagement (impressions, likes, replies) on the right. Choose 1 week, 3 months, 6 months or 1 year, and the previous period overlays as a dashed line and shows the difference.
Below that is a table of impressions, views, likes, reposts and replies for each post you published in the last 30 days, which you can sort by engagement or by impressions. You can also see average engagement per post by day of week and time of day, and the difference in engagement with and without images or video. The numbers can be exported to CSV, so you can tally them in a spreadsheet in your own format.
Numbers are read once a day. Posts you published are read every day until day 3, and once more on day 7. When you want the latest numbers quickly, press "Refresh". Threads, LinkedIn and TikTok are not counted for now, because the networks have not granted permission to read engagement. See pricing for plans that include it, and View analytics for how to use it.
Dùng thử miễn phí 7 ngày. Không cần thẻ tín dụng.
Bắt đầu miễn phíFrequently asked questions
Should I do analytics every day?
Daily numbers vary a lot and are hard to base decisions on. Looking once a month, then deciding on one thing to change, does more for your next posts. Look daily only during a campaign or when you sense a crisis building.
What numbers count as "good"?
It differs greatly by industry, follower count and network, so comparing with other companies' numbers or general benchmarks will not tell you. Compare against your own account last month, and your own posts against each other.
Each network names its numbers differently and I get confused
Fit them into three groups: impressions, reach and engagement. Even the same word, "engagement", includes different actions on different networks, so when you line up numbers across networks, check what is being counted before comparing.
Is analytics pointless while I have few followers?
No. Even with few followers you can see which posts reached people outside your audience and which drew reactions. When numbers are small, differences may not be clear, so we recommend looking at a month at a time.
Any tips for showing analytics to my boss or clients?
Do not just line up numbers. Add one or two sentences on "what happened" and "what we'll do next." Using the same format every month makes it easier to compare with the previous month.
How long should I keep the numbers?
Keep at least a year so you can compare with the same month last year. A network's analytics screen may only let you look back a limited time, so export to CSV every month and store it in a set place inside the company.
How do I see how posts relate to sales?
Social media numbers alone cannot show their relationship with sales. Combine them with numbers from outside, for example by tagging links and counting on your website, or by asking customers who visited how they heard about you. Even lining up each month's post engagement and number of inquiries gives you a clue to the relationship.
