WEBVTT

00:00:00.010 --> 00:00:04.950
You're not imagining it search results did get

00:00:04.950 --> 00:00:09.710
worse Try this sometime search for a product

00:00:09.710 --> 00:00:15.589
you generally genuinely want like best router

00:00:15.589 --> 00:00:20.609
for apartment or how to fix a laptop battery

00:00:20.609 --> 00:00:25.210
drain and Count how many results feel like they

00:00:25.210 --> 00:00:28.390
were written by someone who actually touched

00:00:28.390 --> 00:00:33.020
a damn thing Now, here's the uncomfortable truth.

00:00:34.060 --> 00:00:38.159
The web didn't just get bigger, the incentives

00:00:38.159 --> 00:00:43.899
changed. And if we want good search again in

00:00:43.899 --> 00:00:48.679
2026, we have to talk about what search engines

00:00:48.679 --> 00:00:53.060
actually are optimizing for, because it might

00:00:53.060 --> 00:00:57.460
not be what you think. Welcome back to the Tyler

00:00:57.460 --> 00:01:00.899
Woodward Project. I'm Tyler. a broadcast engineer

00:01:00.899 --> 00:01:05.239
by trade, a Linux nerd by choice, and I really

00:01:05.239 --> 00:01:09.840
enjoy demystifying tech that's supposedly too

00:01:09.840 --> 00:01:14.379
complicated for people. Today, I'm answering

00:01:14.379 --> 00:01:21.280
two questions. First, why did Google search results

00:01:21.280 --> 00:01:25.680
feel like they slid downhill, especially over

00:01:25.680 --> 00:01:31.719
the last few years? Second, What would a good

00:01:31.719 --> 00:01:37.000
search engine optimize for in 2026? Now that

00:01:37.000 --> 00:01:41.859
AI answers are sitting right under the top of

00:01:41.859 --> 00:01:46.420
the page now. I'm going to keep this practical.

00:01:47.019 --> 00:01:51.099
I'll explain the mechanics once and plain English.

00:01:51.700 --> 00:01:55.239
Then we'll talk about incentives and then I'll

00:01:55.239 --> 00:01:59.909
give you a set of tactics. you can use immediately

00:01:59.909 --> 00:02:03.450
to get better results without needing a computer

00:02:03.450 --> 00:02:08.909
science degree or a ritual sacrifice to the algorithms.

00:02:10.270 --> 00:02:15.330
One quick glossary up front. SEO. Which means

00:02:15.330 --> 00:02:19.810
Search Engine Optimization. Basically the art

00:02:19.810 --> 00:02:26.789
and science of getting a page to rank. SERP.

00:02:27.769 --> 00:02:32.870
SERP means search engine results page, the page

00:02:32.870 --> 00:02:38.509
you see after you search. And when I say AI answers,

00:02:38.750 --> 00:02:43.229
I mean the search engine generated generation

00:02:43.229 --> 00:02:47.949
of summary often using a technique called retrieval

00:02:47.949 --> 00:02:52.830
augmented generation or RAG, R -A -G. It retrieves

00:02:52.830 --> 00:02:55.810
some sources and then generates an answer from

00:02:55.810 --> 00:03:01.689
them or tries to. Alright, let's talk about why

00:03:01.689 --> 00:03:12.550
search feels worse. At a high level, Google's

00:03:12.550 --> 00:03:15.810
job is to match your search to the most useful

00:03:15.810 --> 00:03:18.689
stuff on the web. That sounds simple enough,

00:03:18.830 --> 00:03:22.830
right? But it's a hostile environment. Spammers

00:03:22.830 --> 00:03:27.189
want traffic. Publishers want revenue. And Google

00:03:27.189 --> 00:03:30.330
wants to keep you searching while also making

00:03:30.330 --> 00:03:33.770
a little money. So the ranking system becomes

00:03:33.770 --> 00:03:39.909
an optimization problem. If you reward the wrong

00:03:39.909 --> 00:03:44.810
thing, you get the wrong behavior at internet

00:03:44.810 --> 00:03:49.009
scale. One big shift is monetization pressure.

00:03:49.590 --> 00:03:55.210
Search is a massive business and ads are a massive

00:03:55.210 --> 00:03:59.550
part of it. Even when the ad labeling is technically

00:03:59.550 --> 00:04:03.909
present, the practical effect is that commercial

00:04:03.909 --> 00:04:07.430
searches to anything that smells like I might

00:04:07.430 --> 00:04:12.590
buy something. Get a SERP that's crowded with

00:04:12.590 --> 00:04:16.610
ad units, shopping modules, top picks, and other

00:04:16.610 --> 00:04:21.350
widgets before you ever reach what people traditionally

00:04:21.350 --> 00:04:26.600
think of as the results. And that changes what

00:04:26.600 --> 00:04:31.800
you experience as quality. The first screen becomes

00:04:31.800 --> 00:04:35.579
a product page for Google, not a directory of

00:04:35.579 --> 00:04:41.819
the web. Second shift, SEO evolved from help

00:04:41.819 --> 00:04:49.519
search engines understand my page into an adversarial

00:04:49.519 --> 00:04:54.870
industry. If ranking is valuable, People will

00:04:54.870 --> 00:05:00.730
game it if the game changes they'll adapt And

00:05:00.730 --> 00:05:05.449
now we have an ecosystem full of content designed

00:05:05.449 --> 00:05:11.009
to rank first and help later if at all a lot

00:05:11.009 --> 00:05:20.709
of this is affiliate arbitrage Pages optimized

00:05:20.709 --> 00:05:26.420
to catch a query like Best standing desk. Funnel

00:05:26.420 --> 00:05:30.740
you through affiliate links and collect a commission.

00:05:31.920 --> 00:05:34.920
Some of those pages are genuinely useful, but

00:05:34.920 --> 00:05:37.959
many are just rewritten summaries of other summaries

00:05:37.959 --> 00:05:41.779
padded with expert phrasing and engineered to

00:05:41.779 --> 00:05:48.259
look authoritative. Third shift. The measurement

00:05:48.259 --> 00:05:52.980
problem. Search engines can't directly measure

00:05:52.980 --> 00:05:58.040
did this help the person they approximate with

00:05:58.040 --> 00:06:01.439
signals like click -through rate dwell time and

00:06:01.439 --> 00:06:04.800
whether you come back and search again those

00:06:04.800 --> 00:06:12.319
signals are easy to misread if you click a result

00:06:12.319 --> 00:06:16.319
spend let's say three minutes on it because it's

00:06:16.319 --> 00:06:19.160
confusing as hell and then come back and search

00:06:19.160 --> 00:06:26.110
again is that Good outcome or a bad one The system

00:06:26.110 --> 00:06:31.509
all it sees is activity You experienced frustration

00:06:31.509 --> 00:06:39.389
Fourth shift content volume exploded and not

00:06:39.389 --> 00:06:46.529
just from people AI tools. Yeah made it cheap

00:06:46.529 --> 00:06:51.689
to produce plausible looking text at a ridiculous

00:06:51.689 --> 00:06:57.529
scale. Even before modern generative AI, content

00:06:57.529 --> 00:07:01.110
farms could churn. Now it's worse because the

00:07:01.110 --> 00:07:04.009
writing looks smoother while still being empty.

00:07:04.430 --> 00:07:11.069
Or worse, confidently wrong. This creates a pollution

00:07:11.069 --> 00:07:15.670
problem. The web is full of pages that look like

00:07:15.670 --> 00:07:19.850
answers. Search engines have to filter hard.

00:07:20.120 --> 00:07:23.560
And any filtering mistake is now amplified because

00:07:23.560 --> 00:07:28.939
there's so much junk competing for the same queries.

00:07:30.759 --> 00:07:35.319
Fifth, shift. The open web got a whole lot thinner.

00:07:36.839 --> 00:07:40.480
A lot of the best, how do I fix this weird thing?

00:07:42.220 --> 00:07:45.040
That knowledge moved into places that are hard

00:07:45.040 --> 00:07:49.540
to index well, or hard to access at all. Community

00:07:49.540 --> 00:07:53.060
platforms, walled gardens, private groups, chat

00:07:53.060 --> 00:07:56.420
servers, paid walled sites, apps and tools that

00:07:56.420 --> 00:08:00.259
don't expose much to crawlers. Sometimes this

00:08:00.259 --> 00:08:02.879
knowledge exists, but it's not easily visible

00:08:02.879 --> 00:08:08.040
to classic search. So the algorithm overweighs

00:08:08.040 --> 00:08:13.759
what it can see. SEOed pages, aggregators, and

00:08:13.759 --> 00:08:20.579
big domains that publish a lot. 6. Shift. Google

00:08:20.579 --> 00:08:24.060
started answering more queries directly on the

00:08:24.060 --> 00:08:28.519
SERP. That includes featured snippets, knowledge

00:08:28.519 --> 00:08:33.740
panels, people also ask, and now AI generative

00:08:33.740 --> 00:08:38.059
summaries. If you get your answer without clicking

00:08:38.059 --> 00:08:42.279
anything, that's called a zero click. From a

00:08:42.279 --> 00:08:45.500
user perspective, zero click can be convenient.

00:08:45.960 --> 00:08:49.679
From the web ecosystem perspective it reduces

00:08:49.679 --> 00:08:53.279
traffic to publishers and when publishers lose

00:08:53.279 --> 00:08:58.419
traffic they either monetize harder go beyond

00:08:58.419 --> 00:09:02.519
behind paywalls or Optimize more aggressively

00:09:02.519 --> 00:09:05.919
for whatever scraps or you know scraps of traffic

00:09:05.919 --> 00:09:09.399
remain often by targeting queries that still

00:09:09.399 --> 00:09:14.539
trigger clicks That feedback loop tends to make

00:09:14.730 --> 00:09:18.629
The remaining clickable web results, more spammy,

00:09:18.809 --> 00:09:22.309
more affiliate heavy, more extreme in tone, because

00:09:22.309 --> 00:09:28.590
extremes, unfortunately, convert better. This

00:09:28.590 --> 00:09:31.629
is the part that feels very broadcast engineering

00:09:31.629 --> 00:09:35.710
to me. In broadcast, you can't cheat physics.

00:09:36.429 --> 00:09:40.009
When noise rises, you can crank up the gain,

00:09:40.360 --> 00:09:44.200
But you're just amplifying noise. Search is fighting

00:09:44.200 --> 00:09:49.240
the same battle. When the web's signal to noise

00:09:49.240 --> 00:09:52.200
ratio drops, the ranking system can get more

00:09:52.200 --> 00:09:57.480
complex. But complexity doesn't magically create

00:09:57.480 --> 00:10:00.659
signal. You still have to find clean sources

00:10:00.659 --> 00:10:04.460
or you're just making higher resolution garbage.

00:10:06.539 --> 00:10:11.649
Now, to be fair, Google has fought spam for decades

00:10:11.649 --> 00:10:15.049
and keep shipping updates that try to reward

00:10:15.049 --> 00:10:19.110
experience credibility and helpfulness, but spammers

00:10:19.110 --> 00:10:23.570
react fast. AI made that reaction time even faster.

00:10:23.950 --> 00:10:28.230
It's a cat and mouse game. So why does it matter?

00:10:28.909 --> 00:10:32.029
Because search is infrastructure. It's how people

00:10:32.029 --> 00:10:34.690
troubleshoot their tech, choose products, learn

00:10:34.690 --> 00:10:38.820
skills and verify claims. When search degrades,

00:10:39.620 --> 00:10:42.820
the cost isn't just annoyance, it's wasted time,

00:10:43.259 --> 00:10:45.919
worse decisions, and people getting nudged toward

00:10:45.919 --> 00:10:50.360
whatever's most profitable, not what's most accurate.

00:10:52.860 --> 00:10:57.659
Next, I want to make this real with a concrete

00:10:57.659 --> 00:11:01.879
story. What the bad search experience looks like

00:11:01.879 --> 00:11:07.120
now and how AI answers changed to failure modes.

00:11:09.240 --> 00:11:13.320
Let's run a typical modern search query. Say

00:11:13.320 --> 00:11:19.080
you search best budget laser printer 2026 or

00:11:19.080 --> 00:11:22.559
best wireless earbuds for calls. Here's what

00:11:22.559 --> 00:11:26.720
tends to happen. First, you know, the first screen

00:11:26.720 --> 00:11:29.559
shopping modules, sponsored placements and a

00:11:29.559 --> 00:11:33.460
few big name sites. Some are legitimate reviews,

00:11:33.519 --> 00:11:37.779
but many are listicles that all recommend the

00:11:37.779 --> 00:11:41.460
same handful of products, because affiliate programs

00:11:41.460 --> 00:11:46.299
and ad partnerships quietly shape what gets covered.

00:11:48.639 --> 00:11:52.500
Then you click one. You get a long page with

00:11:52.500 --> 00:11:56.860
a table of top picks, a bunch of filler paragraphs,

00:11:57.679 --> 00:12:01.039
and suspicions suspiciously similar phrasing

00:12:01.039 --> 00:12:04.320
across multiple sites. You scroll, you skim,

00:12:04.320 --> 00:12:06.340
and you realize you didn't learn anything you

00:12:06.340 --> 00:12:08.960
could have just guessed from the product page.

00:12:10.879 --> 00:12:16.340
So you refine the search, add Reddit, add forum,

00:12:17.220 --> 00:12:22.480
add a model number, add site this filters. Suddenly

00:12:22.480 --> 00:12:25.120
you find a thread where someone says I've owned

00:12:25.120 --> 00:12:28.460
this for 18 months, the rollers wear out, and

00:12:28.460 --> 00:12:32.000
the driver crashes on macOS after the latest

00:12:32.000 --> 00:12:36.320
update. That's what you wanted! Not a top 10,

00:12:36.899 --> 00:12:40.659
but an actual lived experience in specific failure

00:12:40.659 --> 00:12:48.019
modes. Now, layer AI answers on top. In the best

00:12:48.019 --> 00:12:50.899
case, AI gives you a quick summary that points

00:12:50.899 --> 00:12:54.259
you to solid sources, calls out trade -offs,

00:12:54.879 --> 00:12:58.730
and saves you time. In the worst case, it averages

00:12:58.730 --> 00:13:03.389
the internet. If the input sources are polluted,

00:13:03.870 --> 00:13:07.889
affiliate, fluff, outdated pages, scraped content,

00:13:08.370 --> 00:13:12.710
then the AI can produce a confident, cleanly

00:13:12.710 --> 00:13:19.429
written summary of junk. And AI adds a special

00:13:19.429 --> 00:13:28.139
new problem. Providence. With classic search

00:13:28.139 --> 00:13:33.379
you might land on a page and at least know who

00:13:33.379 --> 00:13:38.879
said it when and in what context With AI answers

00:13:38.879 --> 00:13:43.860
that chain can get blurry Even if citations are

00:13:43.860 --> 00:13:49.899
shown they can be incomplete miss Attributed

00:13:49.899 --> 00:13:53.379
or not actually supporting the specific claim

00:13:53.379 --> 00:14:01.059
being made That's not AI. It's not because AI

00:14:01.059 --> 00:14:04.299
is malicious in any way. It's because language

00:14:04.299 --> 00:14:07.360
models are optimized to produce plausible text.

00:14:08.279 --> 00:14:12.259
Without careful grounding, the system can sound

00:14:12.259 --> 00:14:16.620
right while being absolutely wrong. There's also

00:14:16.620 --> 00:14:20.159
a cultural angle here. We're training ourselves

00:14:20.159 --> 00:14:25.879
out of source checking with the SERP. When the

00:14:25.879 --> 00:14:28.879
SERP becomes an answer box, people stop clicking

00:14:28.879 --> 00:14:32.679
through. When people stop clicking through, publishers

00:14:32.679 --> 00:14:36.740
get less incentive to maintain higher effort

00:14:36.740 --> 00:14:40.259
content. When high effort content declines, the

00:14:40.259 --> 00:14:42.899
AI and the ranking systems have less good material

00:14:42.899 --> 00:14:48.220
to learn from. That's a recipe for a knowledge

00:14:48.220 --> 00:14:51.279
ecosystem that slowly replaces primary sources

00:14:51.279 --> 00:14:56.450
with summaries of summaries. Why this matters

00:14:56.450 --> 00:15:00.570
is bigger than shopping. The same mechanics apply

00:15:00.570 --> 00:15:03.629
to health queries, finance questions, and technical

00:15:03.629 --> 00:15:07.529
troubleshooting. Bad search makes it harder to

00:15:07.529 --> 00:15:12.009
verify anything. And that pushes people toward

00:15:12.009 --> 00:15:16.250
either cynicism, nothing's reliable, or blind

00:15:16.250 --> 00:15:19.090
trust. The answer box said so, so it must be

00:15:19.090 --> 00:15:24.799
correct. If we agree the problem is incentives

00:15:24.799 --> 00:15:28.480
plus pollution plus weird measurement then the

00:15:28.480 --> 00:15:32.360
fix can't be bring back the old Google because

00:15:32.360 --> 00:15:39.120
well the web itself has changed. So what would

00:15:39.120 --> 00:15:43.580
a genuinely good search engine optimize for 2026

00:15:43.580 --> 00:15:48.600
even look like? Let's start with the big idea.

00:15:48.960 --> 00:15:54.639
A search engine is only as good as its objective

00:15:54.639 --> 00:16:00.940
function, the thing it tries to maximize. If

00:16:00.940 --> 00:16:06.779
you optimize for more queries, you can accidentally

00:16:06.779 --> 00:16:10.899
reward confusion. If you optimize for more clicks,

00:16:11.919 --> 00:16:15.019
you can accidentally reward clickbait. If you

00:16:15.019 --> 00:16:18.600
optimize for more ad revenue, you will eventually

00:16:18.600 --> 00:16:22.419
get a SERP that feels like a shopping mall with

00:16:22.419 --> 00:16:29.679
a library in the back. So here's my 2026 scorecard

00:16:29.679 --> 00:16:36.059
for good search in plain terms. A good search

00:16:36.059 --> 00:16:41.000
engine should optimize for successful task completion.

00:16:42.179 --> 00:16:46.019
Did you fix the issue? Choose the right product,

00:16:46.419 --> 00:16:49.759
learn the thing, or find the page you meant to

00:16:49.759 --> 00:16:53.860
find. That means measuring satisfaction carefully,

00:16:54.159 --> 00:17:00.320
not just measuring activity. A good search engine

00:17:00.320 --> 00:17:03.980
should optimize for long -term trust, not, did

00:17:03.980 --> 00:17:08.900
this answer feel good in the moment? But when

00:17:08.900 --> 00:17:13.799
people check later, was it right? That implies

00:17:13.799 --> 00:17:17.150
stronger feedback loops based on corrections,

00:17:17.750 --> 00:17:20.170
expert review where appropriate, and weighing

00:17:20.170 --> 00:17:25.609
sources with real track records. A good search

00:17:25.609 --> 00:17:30.250
engine should optimize for source transparency.

00:17:31.769 --> 00:17:37.710
If you show an AI answer, it should be easy to

00:17:37.710 --> 00:17:42.349
see what it relied on, what's uncertain, and

00:17:42.349 --> 00:17:47.859
what's being inferred. Here are the receipts,

00:17:48.140 --> 00:17:50.960
should be the first class feature, not a footnote.

00:17:51.859 --> 00:17:54.220
A good search engine should optimize for spam

00:17:54.220 --> 00:17:57.700
resistance as the core product feature. That

00:17:57.700 --> 00:18:03.799
means treating SEO spam like malware. Detect

00:18:03.799 --> 00:18:08.220
patterns, penalize networks, and make it expensive

00:18:08.220 --> 00:18:14.420
to scale junk. This is especially important with

00:18:14.420 --> 00:18:18.609
AI generated content. Because cheap to produce

00:18:18.609 --> 00:18:22.410
changes the economics of spam. A good search

00:18:22.410 --> 00:18:25.930
engine should optimize for diversity of sources

00:18:25.930 --> 00:18:30.049
and formats, not diversity as a buzzword. Diversity

00:18:30.049 --> 00:18:36.490
as resilience. If every query funnels you to

00:18:36.490 --> 00:18:40.329
the same five mega websites, you end up with

00:18:40.329 --> 00:18:48.619
monoculture. And monocultures fail. Now, practical

00:18:48.619 --> 00:18:51.940
takeaways you can use today, even if Google doesn't

00:18:51.940 --> 00:18:55.779
change anything tomorrow. Use search operators

00:18:55.779 --> 00:19:02.539
intentionally. Try site, colon, whatever, quotes

00:19:02.539 --> 00:19:08.799
for exact phrases and exclusions with dash to

00:19:08.799 --> 00:19:12.720
remove garbage domains or terms. If you're troubleshooting,

00:19:13.200 --> 00:19:18.140
add a specific error code, model number, or maybe

00:19:18.140 --> 00:19:24.279
a log snippet in quotes. Bias toward primary

00:19:24.279 --> 00:19:29.160
sources. For software and security topics, look

00:19:29.160 --> 00:19:33.180
for vendor docs, official changelogs, standard

00:19:33.180 --> 00:19:38.440
docs, and reputable communities. For products,

00:19:39.119 --> 00:19:42.859
look for teardown -style reviews. Long -term

00:19:42.859 --> 00:19:45.460
ownership updates and forms where people discuss

00:19:45.460 --> 00:19:50.200
the actual failures Treat AI answers like a draft

00:19:50.200 --> 00:19:56.359
not a verdict use them to get vocabulary and

00:19:56.359 --> 00:20:00.180
Maybe a starting plan then verify with at least

00:20:00.180 --> 00:20:04.220
two solid sources if the AI Can't give you citations

00:20:04.220 --> 00:20:08.640
that actually back the key claim Assume it shaky

00:20:08.640 --> 00:20:13.630
don't trust it Build your own trusted web. Bookmark

00:20:13.630 --> 00:20:16.349
a few sites and communities that consistently

00:20:16.349 --> 00:20:19.890
deliver. Subscribe to RSS feeds where you can

00:20:19.890 --> 00:20:23.910
and keep a small notes file of places that were

00:20:23.910 --> 00:20:27.529
right last time. This sounds old school because

00:20:27.529 --> 00:20:33.029
it is, but it still works. I can't tell you.

00:20:33.079 --> 00:20:35.960
how many bookmarks I have of places I found that

00:20:35.960 --> 00:20:38.559
have fixed problems, and I have found the answer

00:20:38.559 --> 00:20:40.700
I was looking for so I could come back again

00:20:40.700 --> 00:20:47.200
if I needed to. Why this matters, you can't control

00:20:47.200 --> 00:20:51.220
Google's incentives, but you can control your

00:20:51.220 --> 00:20:55.000
workflow. A small shift in how you search can

00:20:55.000 --> 00:20:58.880
save, I don't know, hours, and it reduces the

00:20:58.880 --> 00:21:02.880
change you'll make, you know, It reduces the

00:21:02.880 --> 00:21:06.599
chance you'll make a decision based on whatever

00:21:06.599 --> 00:21:10.140
content was most aggressively optimized for this

00:21:10.140 --> 00:21:15.599
week. So yeah, search got worse. Not because

00:21:15.599 --> 00:21:17.799
engineers forget how to build a search engine,

00:21:17.859 --> 00:21:21.859
but because the web got a whole lot noisier and

00:21:21.859 --> 00:21:25.980
the incentives tilted toward monetization and

00:21:25.980 --> 00:21:31.779
scalable manipulation. The good news is We can

00:21:31.779 --> 00:21:36.220
also define what good should mean in 2026. Success,

00:21:36.420 --> 00:21:39.839
trust, transparency and resistance is spam, especially

00:21:39.839 --> 00:21:44.839
when AI is generating the first thing you see.

00:21:46.440 --> 00:21:49.660
Visit TylerWoodward .me, follow at TylerWoodward

00:21:49.660 --> 00:21:52.220
.me on Instagram and threads, subscribe and like

00:21:52.220 --> 00:21:55.440
the show on your favorite podcast platform. I'll

00:21:55.440 --> 00:21:56.319
catch you next week.