not the one one who feels that YouTube sponsor segments have grow to be longer and extra frequent just lately. Generally, I watch movies that appear to be making an attempt to promote me one thing each couple of seconds.
, it positive is annoying to be bombarded by advertisements.
On this weblog submit, I’ll discover these sponsor segments, utilizing knowledge from a preferred browser extension referred to as SponsorBlock, to determine if the perceived enhance in advertisements truly did occur and in addition to quantify what number of advertisements I’m watching.
I’ll stroll you thru my evaluation, offering code snippets in Sql, DuckDB, and pandas. All of the code is on the market on my GitHub, and because the dataset is open, I may also educate you how one can obtain it, so that you could comply with alongside and play with the information your self.
These are the questions I might be making an attempt to reply on this evaluation:
- Have sponsor segments elevated through the years?
- Which channels have the best proportion of sponsor time per video?
- What’s the density of sponsor segments all through a video?
To get to those solutions, we should cowl a lot floor. That is the agenda for this submit:
Let’s get this began!
SponsorBlock is an extension that permits you to skip advert segments in movies, much like the way you skip Netflix intros. It’s extremely correct, as I don’t bear in mind seeing one improper phase since I began utilizing it round a month in the past, and I watch a whole lot of smaller non-English creators.
You is likely to be asking your self how the extension is aware of which elements of the video are sponsors, and, consider it or not, the reply is thru crowdsourcing!
Customers submit the timestamps for the advert segments, and different customers vote if it’s correct or not. For the typical person, who isn’t contributing in any respect, the one factor you must do is to press Enter to skip the advert.
Okay, now that you realize what SponsorBlock is, let’s speak in regards to the knowledge.
Cleansing the Knowledge
If you wish to comply with alongside, you’ll be able to obtain a replica of the information utilizing this SponsorBlock Mirror (it would take you fairly a couple of minutes to obtain all of it). The database schema may be seen right here, though most of it received’t be helpful for this mission.
As one may anticipate, their database schema is made for the extension to work correctly, and never for some man to mainly leech from an enormous group effort to search out what proportion of advertisements his favourite creator runs. For this, some work will have to be accomplished to scrub and mannequin the information.
The one two tables which are necessary for this evaluation are:
sponsorTimes.csv: That is crucial desk, containing thestartTimeandendTimeof all crowdsourced sponsor segments. The CSV is round 5GB.videoInfo.csv: Accommodates the video title, publication date, and channel ID related to every video.
Earlier than we get into it, these are all of the libraries I ended up utilizing. I’ll clarify the much less apparent ones as we go.
pandas
duckdb
requests
requests-cache
python-dotenv
seaborn
matplotlib
numpyStep one, then, is to load the information. Surprisingly, this was already a bit difficult, as I used to be getting a whole lot of errors parsing some rows of the CSV. These had been the settings I discovered to work for almost all of the rows:
import duckdb
import os
# Connect with an in-memory DuckDB occasion
con = duckdb.join(database=':reminiscence:')
sponsor_times = con.read_csv(
"sb-mirror/sponsorTimes.csv",
header=True,
columns={
"videoID": "VARCHAR",
"startTime": "DOUBLE",
"endTime": "DOUBLE",
"votes": "INTEGER",
"locked": "INTEGER",
"incorrectVotes": "INTEGER",
"UUID": "VARCHAR",
"userID": "VARCHAR",
"timeSubmitted": "DOUBLE",
"views": "INTEGER",
"class": "VARCHAR",
"actionType": "VARCHAR",
"service": "VARCHAR",
"videoDuration": "DOUBLE",
"hidden": "INTEGER",
"status": "DOUBLE",
"shadowHidden": "INTEGER",
"hashedVideoID": "VARCHAR",
"userAgent": "VARCHAR",
"description": "VARCHAR",
},
ignore_errors=True,
quotechar="",
)
video_info = con.read_csv(
"sb-mirror/videoInfo.csv",
header=True,
columns={
"videoID": "VARCHAR",
"channelID": "VARCHAR",
"title": "VARCHAR",
"revealed": "DOUBLE",
},
ignore_errors=True,
quotechar=None,
)
# Ignore warnings
import warnings
warnings.filterwarnings('ignore')Here’s what a pattern of the information appears to be like like:
con.sql("SELECT videoID, startTime, endTime, votes, locked, class FROM sponsor_times LIMIT 5")
con.sql("SELECT * FROM video_info LIMIT 5")
Understanding the information within the sponsorTimes desk is ridiculously necessary, in any other case, the cleansing course of received’t make any sense.
Every row represents a user-submitted timestamp for a sponsored phase. Since a number of customers can submit segments for a similar video, the dataset incorporates duplicate and probably incorrect entries, which is able to have to be handled throughout cleansing.
To seek out incorrect segments, I’ll use the votes and the locked column, because the latter one represents segments that had been confirmed to be right.
One other necessary column is the class. There are a bunch of classes like Intro, Outro, Filler, and so forth. For this evaluation, I’ll solely work with Sponsor and Self-Promo.
I began by making use of some filters:
CREATE TABLE filtered AS
SELECT
*
FROM sponsor_times
WHERE class IN ('sponsor', 'selfpromo') AND (votes > 0 OR locked=1)Filtering for locked segments or segments with greater than 0 votes was an enormous resolution. This decreased the dataset by an enormous proportion, however doing so made the information very dependable. For instance, earlier than doing this, all the Prime 50 channels with the best proportion of advertisements had been simply spam, random channels that ran 99.9% of advertisements.
With this accomplished, the subsequent step is to get a dataset the place every sponsor phase reveals up solely as soon as. For instance, a video with a sponsor phase initially and one other on the finish ought to have solely two rows of information.
That is very a lot not the case to this point, since in a single video we will have a number of user-submitted entries for every phase. To do that, I’ll use window capabilities to determine if two or extra rows of information signify the identical phase.
The primary window operate compares the startTime of 1 row with the endTime of the earlier. If these values don’t overlap, it means they’re entries for separate segments, in any other case they’re repeated entries for a similar phase.
CREATE TABLE new_segments AS
SELECT
-- Coalesce to TRUE to take care of the primary row of each window
-- because the values are NULL, however it ought to depend as a brand new phase.
COALESCE(startTime > LAG(endTime)
OVER (PARTITION BY videoID ORDER BY startTime), true)
AS new_ad_segment,
*
FROM filtered
The new_ad_segment column is TRUE each time a row represents a brand new phase of a video. The primary two rows, as their timestamps overlap, are correctly marked as the identical phase.
Subsequent up, the second window operate will label every advert phase by quantity:
CREATE TABLE ad_segments AS
SELECT
SUM(new_ad_segment)
OVER (PARTITION BY videoID ORDER BY startTime)
AS ad_segment,
*
FROM new_segments
Lastly, now that every phase is correctly numbered, it’s simple to get the phase that’s both locked or has the best quantity of votes.
CREATE TABLE unique_segments AS
SELECT DISTINCT ON (videoID, ad_segment)
*
FROM ad_segments
ORDER BY videoID, ad_segment, locked DESC, votes DESC
That’s it! Now this desk has one row for every distinctive advert phase, and I can begin exploring the information.
If these queries really feel sophisticated, and also you want a refresher on window capabilities, try this weblog submit that can educate you all it’s essential to find out about them! The final instance coated within the weblog submit is sort of precisely the method I used right here.
Exploring and Enhancing the Knowledge
Lastly, the dataset is nice sufficient to begin exploring. The very first thing I did was to get a way of the dimensions of the information:
- 36.0k Distinctive Channels
- 552.6k Distinctive Movies
- 673.8k Distinctive Sponsor Segments, for a mean of 1.22 segments per video
As talked about earlier, filtering by segments that had been both locked or had not less than 1 upvote, decreased the dataset massively, by round 80%. However that is the worth I needed to pay to have knowledge that I might work with.
To examine if there may be nothing instantly improper with the information, I gathered the channels which have essentially the most quantity of movies:
CREATE TABLE top_5_channels AS
SELECT
channelID,
depend(DISTINCT unique_segments.videoID) AS video_count
FROM
unique_segments
LEFT JOIN video_info ON unique_segments.videoID = video_info.videoID
WHERE
channelID IS NOT NULL
-- Some channel IDs are clean
AND channelID != '""'
GROUP BY
channelID
ORDER BY
video_count DESC
LIMIT 5
The quantity of movies per channel appears to be like reasonable… However that is horrible to work with. I don’t need to go to my browser and search for channel IDs each time I need to know the identify of a channel.
To repair this, I created a small script with capabilities to get these values from the YouTube API in Python. I’m utilizing the library requests_cache to verify I received’t be repeating API calls and depleting the API limits.
import requests
import requests_cache
from dotenv import load_dotenv
import os
load_dotenv()
API_KEY = os.getenv("YT_API_KEY")
# Cache responses indefinitely
requests_cache.install_cache("youtube_cache", expire_after=None)
def get_channel_name(channel_id: str) -> str:
url = (
f"https://www.googleapis.com/youtube/v3/channels"
f"?half=snippet&id={channel_id}&key={API_KEY}"
)
response = requests.get(url)
knowledge = response.json()
attempt:
return knowledge.get("objects", [])[0].get("snippet", {}).get("title", "")
besides (IndexError, AttributeError):
return ""In addition to this, I additionally created very related capabilities to get the nation and thumbnail of every channel, which might be helpful later. If you happen to’re within the code, examine the GitHub repo.
On my DuckDB code, I’m now capable of register this Python operate and name them inside SQL! I simply have to be very cautious to at all times use them on aggregated and filtered knowledge, in any other case, I can say bye-bye to my API quota.
# This the script created above
from youtube_api import get_channel_name
# Strive registering the operate, ignore if already exists
attempt:
con.create_function('get_channel_name', get_channel_name, [str], str)
besides Exception as e:
print(f"Skipping operate registration (probably already exists): {e}")
# Get the channel names
channel_names = con.sql("""
choose
channelID,
get_channel_name(channelID) as channel_name,
video_count
from top_5_channels
""")
Significantly better! I seemed up two channels that I’m aware of on YouTube for a fast sanity examine. Linus Tech Suggestions has a complete of seven.2k movies uploaded, with 2.3k current on this dataset. Players Nexus has 3k movies, with 700 within the dataset. Seems to be adequate for me!
The very last thing to do, earlier than shifting over to really answering the query I set myself to reply, is to have an concept of the typical length of movies.
This matches my expectations, for essentially the most half. I’m nonetheless a bit stunned by the quantity of 20-40-minute movies, as for a few years the “meta” was to have movies of 10 minutes to maximise YouTube’s personal advertisements.
Additionally, I assumed these buckets of video durations used within the earlier graph had been fairly consultant of how I take into consideration video lengths, so I might be sticking with them for the subsequent sections.
For reference, that is the pandas code used to create these buckets.
video_lengths = con.sql("""
SELECT DISTINCT ON (videoID)
videoID,
videoDuration
FROM
unique_segments
WHERE
videoID IS NOT NULL
AND videoDuration > 0
"""
).df()
# Outline customized bins, in minutes
bins = [0, 3, 7, 12, 20, 40, 90, 180, 600, 9999999]
labels = ["0-3", "3-7", "7-12", "12-20", "20-40", "40-90", "90-180", "180-600", "600+"]
# Assign every video to a bucket (trasnform length to min)
video_lengths["duration_bucket"] = pd.reduce(video_lengths["videoDuration"] / 60, bins=bins, labels=labels, proper=False)The massive query. This can show if I’m being paranoid or not about everybody making an attempt to promote me one thing always. I’ll begin, although, by answering a less complicated query, which is the share of sponsors for various video durations.
My expectation is that shorter movies have a better share of their runtime from sponsors compared to longer movies. Let’s examine if that is truly the case.
CREATE TABLE video_total_ads AS
SELECT
videoID,
MAX(videoDuration) AS videoDuration,
SUM(endTime - startTime) AS total_ad_duration,
SUM(endTime - startTime) / 60 AS ad_minutes,
SUM(endTime - startTime) / MAX(videoDuration) AS ad_percentage,
MAX(videoDuration) / 60 AS video_duration_minutes
FROM
unique_segments
WHERE
videoDuration > 0
AND videoDuration < 5400
AND videoID IS NOT NULL
GROUP BY
videoIDTo maintain the visualization easy, I’m making use of related buckets, however solely as much as 90 minutes.
# Outline length buckets (in minutes, as much as 90min)
bins = [0, 3, 7, 12, 20, 30, 40, 60, 90]
labels = ["0-3", "3-7", "7-12", "12-20", "20-30", "30-40", "40-60", "60-90"]
video_total_ads = video_total_ads.df()
# Apply the buckets once more
video_total_ads["duration_bucket"] = pd.reduce(video_total_ads["videoDuration"] / 60, bins=bins, labels=labels, proper=False)
# Group by bucket and sum advert instances and complete durations
bucket_data = video_total_ads.groupby("duration_bucket")[["ad_minutes", "videoDuration"]].sum()
# Convert to proportion of complete video time
bucket_data["ad_percentage"] = (bucket_data["ad_minutes"] / (bucket_data["videoDuration"] / 60)) * 100
bucket_data["video_percentage"] = 100 - bucket_data["ad_percentage"]
As anticipated, if you happen to’re watching shorter-form content material on YouTube, then round 10% of it’s sponsored! Movies of 12–20 min in length have 6.5% of sponsors, whereas 20–30 min have solely 4.8%.
To maneuver ahead to the year-by-year evaluation I want to affix the sponsor instances with the videoInfo desk.
CREATE TABLE video_total_ads_joined AS
SELECT
*
FROM
video_total_ads
LEFT JOIN video_info ON video_total_ads.videoID = video_info.videoIDSubsequent, let’s simply examine what number of movies now we have per 12 months:
SELECT
*,
to_timestamp(NULLIF (revealed, 0)) AS published_date,
extract(12 months FROM to_timestamp(NULLIF (revealed, 0))) AS published_year
FROM
video_total_ads
Not good, not good in any respect. I’m not precisely positive why however there are a whole lot of movies that didn’t have the timestamp recorded. Evidently solely in 2021 and 2022 movies had been reliably saved with their revealed date.
I do have some concepts on how I can enhance this dataset with different public knowledge, however it’s a really time-consuming course of and I’ll depart this for a future weblog submit. I don’t intend to accept a solution based mostly on restricted knowledge, however for now, I should make do with what I’ve.
I selected to maintain the evaluation between the years 2018 and 2023, on condition that these years had extra knowledge factors.
# Limiting the years as for these right here I've an honest quantity of information.
start_year = 2018
end_year = 2023
plot_df = (
video_total_ads_joined.df()
.question(f"published_year >= {start_year} and published_year <= {end_year}")
.groupby(["published_year", "duration_bucket"], as_index=False)
[["ad_minutes", "video_duration_minutes"]]
.sum()
)
# Calculate ad_percentage & content_percentage
plot_df["ad_percentage"] = (
plot_df["ad_minutes"] / plot_df["video_duration_minutes"] * 100
)
plot_df["content_percentage"] = 100 - plot_df["ad_percentage"]
There’s a steep enhance in advert proportion, particularly from 2020 to 2021, however afterward, it plateaus, particularly for longer movies. This makes a whole lot of sense since throughout these years on-line commercial grew quite a bit as folks spent an increasing number of time at dwelling.
For shorter movies, there does appear to be a rise from 2022 to 2023. However as the information is restricted, and I don’t have knowledge for 2024, I can’t get a conclusive reply to this.
Subsequent up, let’s transfer into questions that don’t rely on the publishing date, this manner I can work with a bigger portion of the dataset.
This can be a enjoyable one for me, as I’m wondering if the channels I actively watch are those that run essentially the most advertisements.
Persevering with from the desk created beforehand, I can simply group the advert and video quantity by channel:
CREATE TABLE ad_percentage_per_channel AS
SELECT
channelID,
sum(ad_minutes) AS channel_total_ad_minutes,
sum(videoDuration) / 60 AS channel_total_video_minutes
FROM
video_total_ads_joined
GROUP BY
channelIDI made a decision to filter for channels that had not less than half-hour of movies within the knowledge, as a approach of eliminating outliers.
SELECT
channelID,
channel_total_video_minutes,
channel_total_ad_minutes,
channel_ad_percentage
FROM
ad_percentage_per_channel
WHERE
-- At the least half-hour of video
channel_total_video_minutes > 1800
AND channelID IS NOT NULL
ORDER BY
channel_ad_percentage DESC
LIMIT 50As rapidly talked about earlier, I additionally created some capabilities to get the nation and thumbnail of channels. This allowed me to create this visualization.
I’m undecided if this stunned me or not. A few of the channels on this listing I watch very ceaselessly, particularly Gaveta (#31), a Brazilian YouTuber who covers motion pictures and movie enhancing.
I additionally know that each he and Hall Crew (#32) do a whole lot of self-sponsor, selling their very own content material and merchandise, so perhaps that is additionally the case for different channels!
In any case, the information appears good, and the chances appear to match my guide checks and private expertise.
I might like to know if channels that you simply watch had been current on this listing, and if it stunned you or not!
If you wish to see the Prime 150 Creators, subscribe to my free e-newsletter, as I might be publishing the complete listing in addition to extra details about this evaluation in there!
Have you ever ever thought of at which level of the video advertisements work greatest? Individuals most likely simply skip sponsor segments positioned initially, and simply transfer on and shut the video for these positioned on the finish.
From private expertise, I really feel that I’m extra prone to watch an advert if it performs across the center of a video, however I don’t assume that is what creators do usually.
My purpose, then, is to create a heatmap that reveals the density of advertisements throughout a video runtime. Doing this was surprisingly not apparent, and the answer that I discovered was so intelligent that it kinda blew my thoughts. Let me present you.
That is the information wanted for this evaluation. One row per advert, with the timestamp when every phase begins and ends:
Step one is to normalize the intervals, e.g., I don’t care that an advert began at 63s, what I need to know is that if it began at 1% of the video runtime or 50% of the video runtime.
CREATE TABLE ad_intervals AS
SELECT
videoID,
startTime,
endTime,
videoDuration,
startTime / videoDuration AS start_fraction,
endTime / videoDuration AS end_fraction
FROM
unique_segments
WHERE
-- Simply to verify we do not have unhealthy knowledge
videoID IS NOT NULL
AND startTime >= 0
AND endTime <= videoDuration
AND startTime < endTime
-- Lower than 40h
AND videoDuration < 144000
Nice, now all intervals are comparable, however the issue is much from solved.
I would like you to assume, how would you clear up this? If I requested you “At 10% runtime out of all movies, what number of advertisements are working?”
I don’t consider that that is an apparent drawback to resolve. My first intuition was to create a bunch of buckets, after which, for every row, I might ask “Is there an advert working at 1% of the runtime? What about at 2%? And so forth…”
This appeared like a horrible concept, although. I wouldn’t be capable of do it in SQL, and the code to resolve it will be extremely messy. Ultimately, the implementation of the answer I discovered was remarkably easy, utilizing the Sweep Line Algorithm, which is an algorithm that’s usually utilized in programming interviews and puzzles.
I’ll present you ways I solved it however don’t fear if you happen to don’t perceive what is going on. I’ll share different sources so that you can study extra about it in a while.
The very first thing to do is to remodel every interval (startTime, endTime) into two occasions, one that can depend as +1 when the advert begins, and one other that can depend as -1 when the advert finishes. Afterward, simply order the dataset by the “begin time”.
CREATE TABLE ad_events AS
WITH unioned as (
-- That is crucial step.
SELECT
videoID,
start_fraction as fraction,
1 as delta
FROM ad_intervals
UNION ALL
SELECT
videoID,
end_fraction as fraction,
-1 as delta
FROM ad_intervals
), ordered AS (
SELECT
videoID,
fraction,
delta
FROM ad_events
ORDER BY fraction, delta
)
SELECT * FROM ordered
Now it’s already a lot simpler to see the trail ahead! All I’ve to do is use a working sum on the delta column, after which, at any level of the dataset, I can know what number of advertisements are working!
For instance, if from 0s to 10s three advertisements began, however two of these additionally completed, I might have a delta of +3 after which -2, which suggests that there’s just one advert at present working!
Going ahead, and to simplify the information a bit, I first around the fractions to 4 decimal factors and combination them. This isn’t mandatory, however having too many rows was an issue when making an attempt to plot the information. Lastly, I divide the quantity of working advertisements by the whole quantity of movies, to have it as a proportion.
CREATE TABLE ad_counter AS
WITH rounded_and_grouped AS (
SELECT
ROUND(fraction, 4) as fraction,
SUM(delta) as delta
FROM ad_events
GROUP BY ROUND(fraction, 4)
ORDER BY fraction
), running_sum AS (
SELECT
fraction,
SUM(delta) OVER (ORDER BY fraction) as ad_counter
FROM rounded_and_grouped
), density AS (
SELECT
fraction,
ad_counter,
ad_counter / (SELECT COUNT(DISTINCT videoID) FROM unique_segments_filtered) as density
FROM running_sum
)
SELECT * FROM density
With this knowledge not solely do I do know that initially of the movies (0.0% fraction), there are 69987 movies working advertisements, this additionally represents 17% of all movies within the dataset.
Now I can lastly plot it as a heatmap:
As anticipated, the bumps on the extremities present that it’s far more widespread for channels to run advertisements initially and finish of the video. It’s additionally fascinating that there’s a plateau across the center of the video, however then a drop, because the second half of the video is mostly extra ad-free.
What I discovered humorous is that it’s apparently widespread for some movies to begin right away with an advert. I couldn’t image this, so I manually checked 10 movies and it’s truly true… I’m undecided how consultant it’s, however a lot of the ones that I opened had been gaming-related and in Russian, and so they began immediately with advertisements!
Earlier than we transfer on to the conclusions, what did you consider the answer to this drawback? I used to be stunned at how easy was doing this with the Sweep Line trick. If you wish to know extra about it, I just lately revealed a weblog submit masking some SQL Patterns, and the final one is strictly this drawback! Simply repackaged within the context of counting concurrent conferences.
Conclusion
I actually loved doing this evaluation because the knowledge feels very private to me, particularly as a result of I’ve been hooked on YouTube these days. I additionally really feel that the solutions I discovered had been fairly passable, not less than for essentially the most half. To complete it off, let’s do a final recap!
Have Sponsor Segments Elevated Over the Years?
There was a transparent enhance from 2020 to 2021. This was an impact that occurred all through all digital media and it’s clearly proven on this knowledge. In newer years, I can’t say whether or not there was a rise or not, as I don’t have sufficient knowledge to be assured.
Which Channels Have the Highest Share of Sponsor Time Per Video?
I acquired to create a really convincing listing of the Prime 50 channels that run the best quantity of advertisements. And I found that a few of my favourite creators are those that spend essentially the most period of time making an attempt to promote me one thing!
What’s the density of sponsor segments all through a video?
As anticipated, most individuals run advertisements initially and the tip of movies. In addition to this, a whole lot of creators run advertisements across the center of the video, making the second half barely extra ad-free.
Additionally, there are YouTubers who instantly begin a video with advertisements, which I believe it’s a loopy technique.
Different Learnings and Subsequent Steps
I appreciated how clear the information was in exhibiting the share of advertisements in numerous video sizes. Now I do know that I’m most likely spending 5–6% of my time on YouTube watching advertisements if I’m not skipping them since I largely watch movies which are 10–20 min.
I’m nonetheless not totally pleased although with the year-by-year evaluation. I’ve already seemed into different knowledge and downloaded greater than 100 GB of YouTube metadata datasets. I’m assured that I can use it, along with the YouTube API, to fill some gaps and get a extra convincing reply to my query.
Visualization Code
You might need seen that I didn’t present snippets to plot the charts proven right here. This was on goal to make the weblog submit extra readable, as matplotlib code occupies a whole lot of house.
You’ll find all of the code in my GitHub repo, that approach you’ll be able to copy my charts if you wish to.
That’s it for this one! I actually hope you loved studying this weblog submit and discovered one thing new!
If you happen to’re interested in fascinating matters that didn’t make it into this submit, or take pleasure in studying about knowledge, subscribe to my free e-newsletter on Substack. I publish each time I’ve one thing genuinely fascinating to share.
Wish to join immediately or have questions? Attain out anytime at mtrentz.com.
All pictures and animations by the writer until acknowledged in any other case.
