This keyword-based advertising is our primary business model.
The rest of the post is a rant about user tracking.
I personally found his argument so persuasive I intend to set everything to DuckDuckGo again now.
Biggest differences to me is that Google seems to handle natural questions like "how do i..." in a "smart" way whereas ddg seems to be more altavista like in that your keywords have higher weight rather than trying to outsmart you. Another difference is that Google is much better on local search thanks to its strong bonds with Google maps, like checking opening hours. For anything not maps related DDG has been my default for can't even remember how many years, over 5.
There are a bunch of these. !a forwards to Amazon, !w to Wikipedia, etc.
But I agree. I check on DuckDuckGo, if I can't find what I'm looking for, most of the time, adding !s finds it. If not, I go back and switching to !g.
It's the opposite for me. I use primarily google, but when the results suck (which is quite often), I switch to duckduckgo to obtain a sensible search.
Often with technical searches the DDG results seem better.
As to rerun the query, that's what the back button is for. Simply hit back, add your bang and enter. Most browsers keep that page in cache so going back is pretty quick.
Sometimes the posts are blatant ads that don't really answer the question, or they add nothing that was not already answered. But sometimes they do answer the question and provide value, while also subtly pointing to a product. I think the latter is acceptable, and I do it occasionally myself. (Example: https://www.quora.com/What-kinds-of-lights-are-best-to-comba...)
Consider this: founders have the strongest conviction for solving a given problem. They often created a product because they saw a gap in the market, and they tend to be domain experts. So, they are highly qualified to answer questions within their area of expertise.
If a post provides value and just so happens to offer a relevant solution with financial strings attached, I have absolutely no problem with it. I might even end up purchasing it, because by searching or asking questions on Quora, I am looking for a solution myself.
edit: And I still do not understand the relation to a search engine :)
The thing is that people (including those who don't use DDG) remember "the goofy-looking duck". DuckDuckGo would be extremely foolish to throw away the mind share they built so far.
Besides, for whatever it's worth, I happen to like the name and logo. I think it's very memorable and has character unlike generic startup name #13.
Almost. There's also:
> "The only other way we’ve found so far, which currently accounts for a much smaller portion of our revenue, is non-tracking affiliate partnerships with Amazon and eBay.
> "When you visit those sites through DuckDuckGo, including when using !bangs, and subsequently make a purchase, we receive a small commission. This mechanism operates anonymously and there is no personally identifiable information exchanged between us and Amazon or eBay.
Retargeting is a popular model for advertising companies, where the retargeting follows you around the internet. Retargeting is much less valuable if you are just a search company.
I understand how DDG has to say this is the best and smartest way to do things because they have to live with their promise of not creeping on their visitors - so they can't use the same analytics that someone like Google has. It's just marketing.
But the ads that go to the top should be the most valuable, not necessarily go to the buyer with the deepest pockets.
I suppose you could say "car" ads cost $x per click, and then let all advertisers rotate, but that's not helpful either. You could measure which one gets clicked the most and have it move to the top, but that can be gamed. If you had any way of knowing how long someone spends on the advertised websites, or base it off of social media reviews or allow consumers with verified accounts to rate ads, that might work better.
I don't know, I just can't trust an ad format where I know the top click just spends the most money.
Why? That's exactly how it works every where else: Radio, TV, Billboards. If two people want the same spot, the person who ponies up more money gets the spot.
It is my understanding that similar to Google, the first couple of links might be ads, but the rest are organic.
I just did a search for a movie I like - To the right is an infobox with a blurb from Wikipedia + links to rotten tomatoes and IMDB. Under it is an amazon ad to buy a copy.
The top 5 links are IMDB, Rotten Tomatoes, a definition of the word, and a trailer.
Seems pretty unobtrusive and on point to me. If I wanted to buy a blu-ray of the movie I'd probably click that ad.
- CPM: you pay per impression (ie how often the ad is seen)
- CPC: cost per click. You only pay when the user clicks.
- CPA: you only pay when the user clicks and does something (eg buys something)
Display advertising (putting ads on third-party websites eg AdSense) is primarily CPM. Search advertising is primarily CPC (or at least this is how Google operates; I assume DDG is similar).
It's important to understand the implications of this.
In a CPC model the system that selects what ads to show you is doing so to maximize revenue. Since they only make revenue when you click a lot of effort is spent in determining the predicted CTR (click through rate) because:
Revenue = # of impressions x pCTR
So it's not just a question of who pays the most for an ad (that's how display advertising works largely) because a $1 CPC with 1% CTR earns more revenue than a $10 CPC with 0.01% CTR.
This is why I have less problem with the CPC search advertising model than pretty much any other model because your interests and the interests of the search engine are largely aligned. You want the results that are most relevant to you and that might be an organic search result or an ad. As long as the ads are clearly display as such, it's fine.
Also with search there is intent. You've searched for something so there's something you want. Who's to say an ad won't be the most relevant result?
The M in CPM is the Roman numeral for 1000. It's an archaic term, and should be replaced with the SI unit for 1000. Though coining the phrase "kilo-impression" would be a stretch.
This has been my main criticism of Google's direction the last 15 years or so. Seems like search results use to populate handsomely with personal sites and blogs run by hobbyists uninterested in extracting a dollar from the visitor. If the search results could return either (a) a link to a book sold on Amazon or (b) some Phd's site discussing the topic it seems likely that (a) wins and (b) gets buried somewhere on page 2.
Why? I'm not using DDG for the advertisements. I'm using them for search.
> I just can't trust an ad format where I know the top click just spends the most money.
This seems orthogonal to whether the actual target is a keyword or user-tracked data. If two firms want to target the same demographic (e.g., Ford and Chevy want to target 20-something Dodge Ram owners who search for 'truck'), something has to decide which advertisements to display on top.
I know a lot of the big-money keywords used to be things like lead-gen arbitrage in the legal and financial sector. If someone is searching for "life insurance quotes" and they see that the top advertiser is some comparison-shopping site paying $50 per click, will they still think of them as an unbiased referral service, or someone who has to bury that $50 cost somewhere?
You end up showing the ad that maximizes something like bid * click_rate + quality_score * quality_value. quality_value can be tweaked to change the balance between short-term revenue and ad quality.
That's the point of a search engine. If you show those as ads, you'd just be duplicating the top search results every time.