Final report for the efficacy of withdrawals and recalls evaluation
Appendix B: Secondary data analysis
Secondary data analysis on social media posts and different FSA and FSS product types.
Table 1: The largest incident categories for FSA and FSS re and post system review project
| FSA: April 2018 to March 2019 | FSA: April 2019 to December 2019 | FSS: April 2018 to March 2019 | FSS: April 2021 to March 2022 |
|---|---|---|---|
| Pathogenic Microorganisms (16%) | Pathogenic Microorganisms (36%) | Allergens (19%) | Allergens (17%) |
| Allergens (13%) | Allergens (43%) | Regulatory Breaches (16%) | Regulatory Breaches (12%) |
| Clandestine Detection (9%) | Foreign Body (19%) | Microbiological (15%) | Microbiological (25%) |
| Poor or Insufficient Controls (8%) | Not Determined / Other (0.4%) | Chemical (8%) | Chemical (20%) |
Figure 7 FSA and FSS product types 2021 to 2022
Figure 2 Where consumers find out about the food recalls
Source Food and You Wave 3 (2021), N = 1,446
Social media data
The FSA and FSS post product withdrawals and recalls information on their dedicated website, Facebook and Twitter pages. Examples of these posts are shown in figures below.
Figure 3 - Examples of FSA and FSS Social Media posts
In 2018/19, the FSA posted approximately 184 posts on their Facebook page, covering topics such as food recalls, food hygiene and food intolerances. In 2021/22, there were approximately 159 posts about food recalls. Table 2 illustrates the number of:
- reach - total number of unique users who viewed posts;
- impressions - how many times content was displayed on a screen; and
- engagement - any action someone takes on the Facebook Page or one of the posts (including likes, shares and comments).
Examples of non-recall posts included food safety for students and Christmas food storage. The majority of the recall-related lifetime post impressions were viewed by people who have previously liked the FSA page, indicating that they have an existing interest in food recalls. It can be seen that the reach for recalls posts was relatively low in 2018/19, especially if compared with the non-recall posts. However, in 2021, the reach for both recalls and non-recalls post was significantly higher indicating better reach of posts to the audience.
The data collected for 2018/19 and 2021/22 varies and it is impossible to compare post impressions and engagements between 2018/19 and 2021/22.
Table 2 – FSA statistics on Facebook
| Recall type | Time period | Post reach | Post impressions | Post engagement |
|---|---|---|---|---|
| Overall average recalls | 2018 to 2019 | 13 | 15 | - |
| Overall average recalls | 2021 to 2022 | 9,719 | - | 332 |
| Overall average non-recalls | 2018 to 2019 | 946 | 1,377 | - |
| Overall average non-recalls | 2021 to 2022 | 10,2777 | - | 458 |
Source: FSA Facebook data analytics
Figure 4: Facebook lifetime post total reach and impressions (2018 to 2019)
Source: FSA Facebook data analytics
Figure 5: Facebook lifetime post total reach and engagements (2018 to 2019)
Source: FSA Facebook data analytics
During 2018/19, the FSA created 193 tweets about food recalls. Like with Facebook posts, Twitter users were less engaged with non-recall tweets than recall-related tweets. For example, non-recalls tweets had an average engagement level of 2.59% (i.e. users liking, replying, retweeting etc. these tweets) compared to an average of 0.79% for non-recalls tweets. According to social media management tools, an engagement level of 0.02% to 0.5% is classified as ‘good’(footnote). However, when analysed by URL clicks, Twitter users were more likely to explore and click the URL with recalls posts than non-recall posts, indicating that these tweets were engaging or interesting for users.
During 2021/22 period, the FSA posted approximately 182 tweets about food recalls. Although the number of tweets remained similar, overall engagement had improved significantly. Average engagement rate was 0.026% which is considered ‘good’. The engagement was up by ten times and the number of URL clicks was up by more nearly 14 times. Overall, these results indicated an improved reach of the FSA tweets.
Table 3 - FSA Twitter reach 2018/19 vs 2021/22
| Recall type | Time period | reach | Post impressions | Post engagement |
|---|---|---|---|---|
| Overall average recalls | 2018 to 2019 | 6,647 | 56 | - |
| Overall average recalls | 2021 to 2022 | 9,719 | - | 332 |
| Overall average non-recalls | 2018 to 2019 | 946 | 1,377 | - |
| Overall average non-recalls | 2021 to 2022 | 10,2777 | - | 458 |
Web analytics
Overall, during 2018/19, there were 1,492,318 unique page views of the 153 alerts on the FSA withdrawals and recalls website. The average time spent on the webpage was around half a minute per user. Although the number of alerts dropped to 68 in 2021/22 with 367,853 unique page views, the average user was spending round 2 minutes on a webpage indicating a significant increase in content engagement.
For the FSS, the website received 46,861 unique page views, with the average user spending 1 minute 12 second on the website in 2018/19. Unfortunately, there was no available data to do the comparison with 2021/22 stats for the FSS.
We also explored the bounce rate, which is the proportion of people that come to a website and leave without clicking to any other pages besides the one they first landed on. On average, the bounce rate for both the FSA and FSS was 68%, suggesting that the information consumers were looking for was easily accessible from a singular page. In 2021/22, the bounce rate for the FSA was down to 63% indicating a slight improvement in the website navigation between pages. Unfortunately, there was no available data to do the comparison with 2021/22 stats for the FSS.
A total of 367 people have subscribed to the FSS email alerts, and 51 to the text alert system. We currently do not have any demographic information on subscribers. Figure 4 below shows new subscriptions (by email and SMS) to the FSA alerts on a monthly basis. A total of 32,963 people subscribed to the FSA alerts from April 2018 to March 2019. Figure 4 below shows a big spike in subscriptions in April 2018 which coincides with the launch of the new system for the alerts.
Figure 6 – New monthly subscriptions to the FSA alerts (2018/19)
Since 2018/19 period, the overall number of platform users continued to grow reaching over 40,000 in 2021/22. At the end of March 2022, there were 41,571 users subscribed to food alerts and 30,867 – to allergy alerts.