Say you open a service to learn a spreadsheet function. You find an explanation, then follow a travel video it offers you. When you close the service, you wonder what happened to your plan.
The travel video might have been a welcome break. It might also have pushed out something you wanted to finish. Calling the whole session “searching” or “scrolling” doesn’t tell you which.
A useful question is whether you’re glad you spent that time on what you watched or read. What you want can include learning, relaxing or finding something you didn’t expect.
This module is for working adults who use services that rank what they show you, when you have something you want to learn, or notice a gap between what you meant to do and what you ended up watching, reading or hearing.
What you’ll get from this module
In this module
By the end of this module, you’ll have a clearer way to judge your choices about what you watch, read and hear: what you wanted, how you found it and whether it was worth it. You’ll learn what to notice, choose a situation from daily life and review one week of choices. An account of how one service worked, an experiment with a made-up social media feed and two studies of real use explain why these questions go together.
Ask Jing
Become a Jingius.
What searching and recommendations each leave you to decide
Company technical account: search can include recommendations.
Carlos A. Gomez-Uribe and Neil Hunt’s 2015 account of Netflix describes search as part of a wider system for recommending things. Your search tells the service what you’re looking for. Its software helps choose the options you see.
Typing a search doesn’t mean you chose everything on the results page. You asked the question; the service helped shape the options. Then you choose among them. Both parts matter when you look back at a session.
Why it matters: This is an account from some years ago, by Netflix’s own engineers. It shows how search and recommendations can work together. It doesn’t measure how much control viewers have.
Read the technical account of Netflix search and recommendations.
Online experiment: looking for a topic helped immediate learning.
Jana H. Dreston and German Neubaum’s 2025 paper looked at news people seek out on purpose and news they come across by chance. Their experiment ran in two rounds. The analysis matched the answers of 921 German university students across both rounds.
The students were told to look for information on one topic in a made-up social media feed. They also came across other topics they hadn’t been told to look for. They knew they would be asked later what they remembered about their topic.
Looking for the topic led to better recall and better recognition than coming across information by chance. Recall meant remembering it with no answers to choose from. Recognition meant picking it out from a set of answers. Coming across information by chance still helped on both, compared with not seeing it at all.
The same feed could hold information students were looking for and information they just came across. Both could teach them something, but looking for a topic made a difference to what they remembered.
Why it matters: The tests came soon after they read the feed. Remembering facts when you’re told what to look for is different from building lasting understanding of a question you chose yourself. The feed was made up; the students weren’t using a real search engine.
Read the experiment on intentional and incidental news consumption.
Observational study: search results and chosen sources differed.
Starting with a question gives you a direction, but the choices you make next still matter. Ronald E. Robertson and colleagues’ 2023 study matched surveys with what people really saw and clicked on in Google Search, around the 2018 and 2020 US elections.
In both rounds, people tended to pick news sources closer to their own politics than the mix their search results showed them. What they picked also included more sources rated unreliable. What people chose was different from what they’d been shown.
This helps explain why starting with a search doesn’t settle whether what you end up reading is good. You can choose actively from the results and still narrow what you come across.
Why it matters: The researchers watched real use. They didn’t run an experiment to pull apart the effects of how people worded searches, the order of the results and why people chose what they did. The finding shows a gap between what was on offer and what people chose.
Read the study comparing Google Search exposure and engagement.
Observational field study: following and recommendations both featured in regret.
Hyunsung Cho and colleagues’ 2021 study looked at how 29 Android users used four social apps. It combined a week of records of app use, short surveys during the day and interviews afterwards.
People said they regretted time spent on parts of the apps that showed posts from accounts they followed. They also described being pulled away from what they’d meant to do by things the app recommended. Following familiar accounts didn’t make a session worth it on its own.
That difference is useful when you judge your own time. A choice you made earlier to follow someone, and what you want today, can be different things. In the same way, the fact that something was recommended doesn’t tell you whether you were glad of it.
Why it matters: These were patterns seen in a small study. They give you a reason to ask what you got out of a session, whether you followed a familiar account or opened something recommended.
Read the field study of regret and smartphone app features.
What this means for you
Keep three questions apart. What did you want before you opened it? How did the material reach you? And was it worth it? Answering only one can hide the gap you want to understand.
Start purposeful learning with a question
Before you open a service to learn something, write down the question you want answered. Then search, or choose a source that fits it. The question helps you judge whether what you find answers what you wanted to understand.
This takes the idea from the made-up feed experiment into everyday learning. Choosing your own question isn’t the same as being given a topic in a study, so notice whether it helps you in your situation.
Keep the question apart from any belief that search results must be good. Robertson’s study shows why what you choose deserves attention as well as what appears. A narrow question can still keep you inside a view you already hold.
Record the origin of a topic separately from its value
For the review below, use four labels. “Pulled” means you chose the topic, the question or the particular thing before the service offered it. “Pushed” means you got interested in the topic after the service offered you something about it.
Use “mixed” when something you already wanted and a topic you were offered both played a part, or when the session moved between them. Use “unclear” when you can’t reliably remember which came first.
These labels describe how a choice started, not whether it was good. Write down separately how the material reached you: through a search, a source you chose, a saved list, a recommendation or a mix.
Go back to the made-up spreadsheet session. The question came first, so that part is pulled. The travel video brought in a new topic, so that part is pushed. Write linked entries, or mark the session mixed.
Now ask whether the travel video was welcome or unwanted. Either answer fits it being pushed. The label alone can’t tell you whether it was worth it.
Say instead you opened a service to relax and find some music. A recommendation picked the artist. That is mixed: what you wanted came first, and the offer gave it a direction.
Relaxing is a real purpose. Choosing to be entertained, or changing your mind, doesn’t have to count as failing. What matters in this review is whether, looking back, you’re glad you made that choice.
Look for one recurring mismatch
Choose which sessions you’ll watch before you start: evening viewing, say, or sessions you open to learn something. Write a short entry close to each one, for seven days. A week gives you a manageable sample to look at.
Note what you wanted before you opened it, and whether the topic came to mind before or after something was offered. Write down how you got to the material, what was useful and whether a change of plan was welcome. Add any drift you didn’t want. If you can’t remember the order, say so. Note any sessions you missed, instead of filling them in from guesses.
At the end, count up all four labels. If you use percentages, divide the number of entries with each label by the number of all your entries, mixed and unclear included. Multiply the result by 100. Each entry counts the same, however long the session lasted. The percentages describe your entries, not time spent or how many things you watched or read.
Then look for a gap that keeps coming back between what you meant to do and what happened. Look for finds that were worth it, too. Choose a change that deals with a gap, instead of trying to raise or lower one label for its own sake.
Keep a queue that still represents what you want
For an interest that keeps coming back, keep a short queue: a list of questions and particular things you chose for a reason. Start sessions about that interest with the list, or with sources you’ve chosen for it.
A saved list helps you get back to something you wanted earlier. But your interests or needs may have changed. Check whether you still want what’s on it, and take off what no longer fits.
Don’t turn a saved list into a duty to get through everything on it. Keep exploring recommendations when finding new things is the point. Judge a find by what it gives you, as well as noticing how it reached you.
Try one change during the next week, and compare how useful the sessions were and how much you drifted. Treat the comparison as a personal look back: other things may also have changed from one week to the next.
How about your daily life?
Where do you notice a gap between opening a service and ending a session? You might choose learning for work, evening viewing or following news about something you care about.
Where does finding new things already work well for you? Keep that in mind when you choose what to look at. One good situation is enough.
Sit down and write some ways you could put this lesson to use in your daily life. What would you like to make easier, understand better or handle differently?
| Situation I want to examine | What I want to understand |
|---|---|
| ____________________ | ____________________ |
| ____________________ | ____________________ |
Which of these would be most useful to work through now?
My chosen situation: ____________________
Choose a situation where you want to learn something, or where what you end up with differs from what you meant to do. Take it into the exercise below. You can change your choice.
Review one week of choices
Start with your chosen situation. If you came straight here, pick a learning situation that keeps coming up, or one where what you end up with differs from what you meant. Include relaxing when that’s what you chose to do.
- Choose your coverage. Name which sessions you’ll write down for seven days, such as evening viewing. Keep track of the ones you miss, so the record describes the use you really saw.
Sessions and dates: ____________________ - Record near the event. Note what you wanted before you opened it, and whether you chose the topic before or after something was offered. Write down how you got to the material and what you got from it. Include a welcome change of plan, or any drift or regret you didn’t want.
Purpose / topic origin / route / result / unwanted drift: ____________________ - Classify without grading. Use pulled for a topic you chose before anything was offered, and pushed for a topic an offer brought in. Use mixed when both played a part or the session changed direction.
Use unclear when you can’t remember the order. You can link two entries for a session that changed. Neither pushed nor mixed means the use was unwanted.
Category / uncertainty / missed sessions: ____________________ - Read the pattern. Count your pulled, pushed, mixed and unclear entries. If you work out percentages, divide each count by all your entries, then multiply by 100. Keep mixed and unclear in the total you divide by; the result describes entries, not time.
Find one gap that keeps coming back, and note welcome finds separately. If you missed too many sessions to see a pattern, write down that gap instead of making up a conclusion.
Counts / coverage limits / mismatch / worthwhile discovery: ____________________ - Choose one relevant change. For learning, write the question down before you look for anything. For an interest that keeps coming back, try a short list of questions or things chosen for a reason, or start with a source you chose.
Keep exploring when you’re glad you did. Check whether saved things still fit; you don’t owe every item your attention. If no unwanted gap shows up, write that down instead of forcing a change.
Change and reason, or reason to keep the current approach: ____________________ - Review the next week. Compare whether the sessions you wrote down did what you wanted, and whether the unwanted drift changed. Keep any gaps in your record visible. This is a personal look back, not a controlled test of what caused a change.
What to keep, change or examine further: ____________________
Ask Jing
Become a Jingius.
Your next step
You finish with a record of the situation, one useful change or a reason to keep your way of doing things, and any gaps still open. Finding new things that are worth it can stay part of that choice.
A note on the evidence
The sources look at different questions: how search sets out choices, what people remember soon after reading, which news they choose and when they regret using apps. Together, they help explain what to look at in your own choices. They don’t show one best route for every service or purpose.
Use your record to understand one situation. The share of entries with a label doesn’t measure how much control a service’s software has over you.
How we prepared this module
This module links a company’s technical account with experiments and studies of real use. Made-up spreadsheet and music examples explain the differences used in the written review. The exercise helps you look at your own purposes and choices.
Sources
- Carlos A. Gomez-Uribe and Neil Hunt. (December 2015). The Netflix Recommender System: Algorithms, Business Value, and Innovation. ACM Transactions on Management Information Systems, 6(4), Article 13. Published technical article; §2.8 Search, p.13:5.
- Jana H. Dreston and German Neubaum. (21 February 2025). Navigating Social Media News Use: Exploring the Impact of Intentional and Incidental News Consumption on Objective and Subjective Political Knowledge. Communication Research, OnlineFirst. Method—Participants and Procedure; Design and Material; Results—Free Recall and Recognition; Limitations and Outlook.
- Ronald E. Robertson, Jon Green, Damian J. Ruck, Katherine Ognyanova, Christo Wilson, and David Lazer. (24 May 2023 online; 8 June 2023 issue). Users choose to engage with more partisan news than they are exposed to on Google Search. Nature, 618, 342–348. Abstract, paragraphs beginning “To address” and “In both waves”.
- Hyunsung Cho, Daeun Choi, Donghwi Kim, Wan Ju Kang, Eun Kyoung Choe, and Sung Ju Lee. (18 October 2021). Reflect, not Regret: Understanding Regretful Smartphone Use with App Feature-Level Analysis. Proceedings of the ACM on Human-Computer Interaction, 5(CSCW2), Article 456. Abstract, study description and findings paragraphs.
