We tracked family reading behaviours through the TANDEM app to assess its impact on improving shared reading
Earlier this year (March 2024), we began working with Early Ideas Limited using their app TANDEM, a story creation app powered by generative AI, alongside parents and caregivers, early-years practitioners and private, public and third-sector organisations. TANDEM's goal is to give all children an equitable start in life by enabling them to experience the early interactions and strong relationships that will set them up for life. The app is designed to assist parents in creating an engaging reading experience with their child at home.
As an innovation partner, we used the TANDEM app as a probe to understand what would be needed to create a high-quality, scalable and low-cost shared reading experience between caregivers and children. Our working hypothesis was that this needed to be a stimulating, fun and tailored experience for low-income families with children under five to use at home and accessible for parents and caregivers across the UK.
This is the fourth update, exploring how we tracked family reading behaviours through the app and created different reading personas based on their habits.
After the previous two test-and-learn cycles, 9 of the 11 research participants opted to participate in a one-month pilot (test-and-learn cycle three), which we designed to track reading behaviours. At the end of the interviews with parents, we assisted them in installing the app on their devices and provided them with ‘dummy accounts’ that were unidentifiable to the app developers. This enabled more privacy for participants while allowing us to track their app usage, such as session start and end times, books generated, session length and total reading time.
We also explained to the parents taking part in the pilot how to capture shared reading moments in a physical diary, which we gave them to use for logging these sessions. We asked them to use the app as and when it felt natural to them so that we could gather a realistic understanding of their usage patterns. The diary included five pages, each representing a week from Monday to Sunday, where participants could track shared reading sessions involving both physical and digital books, noting details such as location, duration, and any barriers to reading.
At the end of each week, we sent a prompt via text and asked participants to share a photo of their logged entries. This enabled us to monitor their activity frequently,so that we could build a picture of each family's reading behaviour on a weekly basis. We were then able to prepare and tailor specific questions for each parent during a 20-minute follow-up call once the pilot ended.
The TANDEM developers also implemented phone notification prompts, so that we could assess whether they would encourage participants to engage with the app at different times of the day. These notifications featured messages when users reached key milestones, such as completing the creation and reading of 1, 3, 5, and 10 stories, and completing a total of 100 pages of stories generated.
At the end of the one-month pilot, we conducted 20-minute telephone interviews with participants, asking tailored questions using a data map (fig 8) we created for each parent to help us further unpack their reading behaviours.
What we learned from the one-month pilot was that the TANDEM app:
Our final research opportunity was a 1 hour 15 minute ‘What if’ workshop, facilitated within Northside Youth Centre, a thriving community hub in Edmonton. We brought together four parents who had participated in the three test-and-learn cycles with the founders of the TANDEM app and key project stakeholders from Nesta.
Together we explored three topics using prompt cards that drew inspiration from The Tarot Cards of Tech. The topics we focused on were:
Attendees were split into two groups to ensure our priority parents were evenly distributed. After 10 minutes of discussion using prompt cards, we reconvened for each group to share their thoughts and hear diverse perspectives from the attendees.
We have synthesised insights from the three test-and-learn cycles (including the one-month pilot) and the ‘What if’ workshop into five different personas or archetypes, which reflect different dimensions of how the families responded to using the TANDEM app.
Before the pilot, a few participants had not previously started or established reading routines with their children, however, two of the caregivers reported a significant change in their daily reading habits using the app.
One caregiver also explained that using TANDEM with her child became a launch pad to explore physical books. The mother highlighted her child’s positive reception to the app, noting that during the pilot she "would always ask me to read a book for her so she can sleep … She is not using the iPad anymore”. This same mother also shared how her child’s vocabulary noticeably improved due to these new reading behaviours.
One father shared that since the pilot he is regularly using the TANDEM app with his son: "Usually in the evenings before bed, it’s the last thing we (now) do”. This father valued the comprehension questions posed by the app after each story, highlighting the impact this had on his son’s engagement with reading.
For some participants, the introduction of TANDEM not only encouraged shared reading but also deepened existing bonds with their children. Even participants who already felt closely connected to their children reported that the app enhanced these relationships. The same father proudly said, "I bond so well with my son, but with TANDEM, it took the bonding to another level." This impact extended beyond structured reading times, with this father elaborating on how his child actively initiated TANDEM sessions with him.
Another participant expressed that TANDEM helped her feel more engaged in her child’s daily activities, sharing, "For me, it changed a lot, to be honest, because now it’s like having a bond … To be honest, before, I never read a storybook for her, but now she’s asking me to read a storybook and have storytime."
Some participants who already maintained consistent reading habits using physical books with their child observed that incorporating digital stories via TANDEM into their routines occasionally disrupted bedtime.
One mother noted that while they enjoyed reading, “Sometimes she [the daughter] wanted two stories, and that was fine, but if she wanted another one, sometimes it would get her a little bit too excited because we mainly used it as a before bed thing”.
Another mother remarked, “Definitely not an evening thing, not a before-bed thing.” The same mother also shared that “If I mentioned it [TANDEM], he would just run over and read straight away and was excited, but he would try to do it all himself.” She went on to say that introducing TANDEM led to some negative behaviours like tantrums, as her son wanted to use it more often than she thought was a good idea.
These participants emphasised that the process of creating and reading a story on TANDEM was noticeably longer than simply reading a physical book, which they found quicker and less likely to overstimulate their children before sleep.
All of the participants' children (2.5-5 year olds) used digital devices on a daily basis. This was not something that we looked for while recruiting, but we did find that digital technology usage was common among children in our priority families.
Two participants whose children regularly engaged in passive tech use throughout the day (eg, watching videos on YouTube) reported positive shifts after integrating TANDEM into their routines. Both were pleasantly surprised by how the app transformed passive screen time into more educational interactions.
One father observed, "He has a tablet he’s so addicted to, but TANDEM shifts his attention from that tablet," while one mother noted, "I don’t give her the phone anymore now … she just wants me to read the book”. This mother shared the positive impact of the app on changing her daughter's digital behaviour to prioritise reading over watching YouTube.
This group of caregivers emerged from discussions during the ‘What if’ workshop, where we intentionally explored this topic. Safety and privacy were top priorities for some parents, who explained that they avoid uploading photos of their child to digital platforms, prefer not to share data with third parties, and favour a low (no)-advert environment.
One participant expressed satisfaction with the current level of personal information the app asked caregivers to share but remained cautious about expanding this further. While they felt comfortable choosing an avatar representing their child’s skin tone, they were reluctant to use real photos of their child on an app. This caregiver highlighted during the workshop that in the event of a data breach, they would be particularly concerned if sensitive details, such as their child’s name or date of birth, were compromised.
In contrast to the more privacy-conscious parent, two other participants exhibited a high level of trust in the app. They were more comfortable with sharing personal data if the purpose was transparent. To gauge their comfort level, we presented scenarios involving using a child’s photo versus an avatar and whether they would be willing to record audio. Both participants indicated that as long as they were informed about the purpose and understood that the features were designed with their child’s benefit in mind, they were open to these options.
One participant said, “I don't mind doing one [photograph]. Yeah, I don't mind anything that can benefit her”, expressing little concern and viewing these features as supportive rather than intrusive. Both of these participants had also witnessed their children learning new words through their exposure to the shared reading app. One of these participants shared, “She started school this year, but she is improving a lot, to be honest, because of TANDEM.”