Use Adwords to find the best title for your Hacker News post(blog.chargeback.cc)
blog.chargeback.cc
Use Adwords to find the best title for your Hacker News post
http://blog.chargeback.cc/2012/11/19/use-adwords-to-find-the-best-title-for-your-hacker-news-post/
22 comments
This is a real problem here, and has been for a while. I think it applies to comments too, quick easy to read and 'obvious' comments earn rapid upvotes, multiparagraph thoughtful pieces get ignored.
Content aggregation sites like HN and reddit often weight upvotes based on time. This makes some sense; an article that gets x upvotes in y minutes is more popular than an article that takes 4*y minutes to get the same number of upvotes!
But on Reddit this has an odd side-effect; as early upvotes are the key to getting to the front page, an article that takes 30 minutes to read doesn't get any upvotes in the first 30 minutes. But a meme or a 10-second gif? Upvoted in seconds, and on its way to the front page.
Of course, Reddit is quite successful; no shame in providing bite-sized posts with quick gratification if that's what the market demands.
But on Reddit this has an odd side-effect; as early upvotes are the key to getting to the front page, an article that takes 30 minutes to read doesn't get any upvotes in the first 30 minutes. But a meme or a 10-second gif? Upvoted in seconds, and on its way to the front page.
Of course, Reddit is quite successful; no shame in providing bite-sized posts with quick gratification if that's what the market demands.
This is a really interesting insight. What do you think would be a good way of levelling the playing field for articles which take 30 minutes to read?
Also, take the title of yesterday's #1 article, and write an article about the opposite
OP - I was aghast at your CTR rates when I first read your post. I don't know your marketing/segment but it's certainly worth questioning whether or not you are getting the right keywords to trigger your ads. Your top three ads have a grand total of 28 clicks over > 21,000 impressions. That is misery. There's just no way you are going to be able to be profitable/break-even with AdWords if you can't improve that CTR.
As such, it leads me to question your keyword selection. You need to know "What were the keywords that actually triggered my ad?" This is possibly VERY different from the keywords you entered (particularly if you are using broad match). My guess is you're wildly off the mark in terms of your keyword selection/filter. Adwords expands broad match keywords to the point of unprofitability IME. You have to find out what people are actually searching for that trigger your ad and "trim the fat" by modifying your keywords so that you don't pay for unrelated things.
For example, let's say you have "payment disputes" as a broad match keyword. That is so general that Google is going to auto-expand your keywords to synonyms and "things that Google thinks are like payments, disputes, and payment disputes". So if you used that as a broad match keyword, you might be surprised to find that your ads were served for the following keyword searches:
* Credit repair atlanta
* bankruptcy dispute
* credit card payment was declined
etc, etc, etc - "similar yet different" is not profitable here. I'd suggest you review the actual keywords and work to get that CTR up.
As such, it leads me to question your keyword selection. You need to know "What were the keywords that actually triggered my ad?" This is possibly VERY different from the keywords you entered (particularly if you are using broad match). My guess is you're wildly off the mark in terms of your keyword selection/filter. Adwords expands broad match keywords to the point of unprofitability IME. You have to find out what people are actually searching for that trigger your ad and "trim the fat" by modifying your keywords so that you don't pay for unrelated things.
For example, let's say you have "payment disputes" as a broad match keyword. That is so general that Google is going to auto-expand your keywords to synonyms and "things that Google thinks are like payments, disputes, and payment disputes". So if you used that as a broad match keyword, you might be surprised to find that your ads were served for the following keyword searches:
* Credit repair atlanta
* bankruptcy dispute
* credit card payment was declined
etc, etc, etc - "similar yet different" is not profitable here. I'd suggest you review the actual keywords and work to get that CTR up.
You learned one of the basic rules of writing good headlines. Don't overstate it. Always get to the point. Quick!
Your first one was really bad.
Show HN: ChargeBack.cc – Battling the last frontier in payments
It has a Star Trek reference, and no mention of what is it that you do. Bad headline. Good for learning.
The next one is worse.
Show HN: ChargeBack.cc – an innovative solution to a messy problem
A messy problem? You know what products relate to messy problems? Diapers, toilet paper, tampons, etc. Really bad headline. Its actually quite funny, though.
The final headline is perfect. Not because of Adwords, but due to the fact that you just gave up on trying to be clever (which is the worst thing you can do on copywriting), and were just honest and straight to the point.
Show HN: ChargeBack.cc – Get your money back
Lets break it down:
Using adwords for writing headlines is clever, but not smart. Because you dont know how effective those ads are (probably not very), and which market it aims to sell to. You are better off reading a good copywriting book.
Your first one was really bad.
Show HN: ChargeBack.cc – Battling the last frontier in payments
It has a Star Trek reference, and no mention of what is it that you do. Bad headline. Good for learning.
The next one is worse.
Show HN: ChargeBack.cc – an innovative solution to a messy problem
A messy problem? You know what products relate to messy problems? Diapers, toilet paper, tampons, etc. Really bad headline. Its actually quite funny, though.
The final headline is perfect. Not because of Adwords, but due to the fact that you just gave up on trying to be clever (which is the worst thing you can do on copywriting), and were just honest and straight to the point.
Show HN: ChargeBack.cc – Get your money back
Lets break it down:
Show HN: - Addressing your audience directly.
Chargeback.cc - Mentioning the product in the headline.
Get your money back - Clearly stating what pain it aims to ease.
Using adwords for writing headlines is clever, but not smart. Because you dont know how effective those ads are (probably not very), and which market it aims to sell to. You are better off reading a good copywriting book.
which copywriting book(s) or material do you recommend?
Tested Advertising Methods by John Caples.
It's an oldie but a goodie. It talks a lot about split testing from the print days -- catalogs, newspaper classifieds, etc. Writing headlines gets its own chapter. Really great book. Really easy to read and understand. The current edition is the 5th edition. You can get it from Amazon.
It's an oldie but a goodie. It talks a lot about split testing from the print days -- catalogs, newspaper classifieds, etc. Writing headlines gets its own chapter. Really great book. Really easy to read and understand. The current edition is the 5th edition. You can get it from Amazon.
I can't recommend http://copyhackers.com/ highly enough. The blog and mailing list are good, the ebooks are excellent. Joanna has figured out how to teach geeks to write powerful copy.
The adweek copywriting book by Joe Sugarman is full of great stuff. You should also buy the cheap fake gossip newspapers in the supermarket and study the ads. You will learn a bunch from them.
I'd recommend Ca$hvertising.
I guess it's somewhat interesting that AdWords can confirm what are good titles and what aren't but I could've told you at the outset that the first two titles are vague meaningless titles that won't draw anyone's attention. "The last frontier in payments"? Really? I know I wouldn't read that because it sounds like marketing BS.
Also, reaching the front page on the weekend has a lower barrier to entry. I see posts with 13 votes hang around for hours.
Also, reaching the front page on the weekend has a lower barrier to entry. I see posts with 13 votes hang around for hours.
It's true they hang around for more but it's also probably harder to gain those 13 votes.
Yup, this is exactly what I've found. Weekends are harder to get votes on, but the posts last much longer once they get traction.
As duiker mentioned, getting those 13 votes is likely quite a bit harder. HN on the weekends is a pretty insufferable place to be, and the activity present seems less like the type to be surfing /new and making quality discussion so much as it appears to be the sort of mob who runs around making pithy snark, and upvoting others who engage in like behavior.
I don't want to criticize, but if getting to the first page is that important to you, I don't know, but it doesn't seems very healthy. Revenues and products are more important than bragging rights.
Ok competition is good, hacking is good, but doing that just for HN sake seems like overengineering for something not important at all (I mean, not central to most business)
However, making a product out of it is something quite different - and it's a great thing you did!
Ok competition is good, hacking is good, but doing that just for HN sake seems like overengineering for something not important at all (I mean, not central to most business)
However, making a product out of it is something quite different - and it's a great thing you did!
I think you underestimate how much attention HN can get you and overestimate the effort of doing what they did here. Seems like a great time investment to me.
I tried similar things to play with the Mechanical Turk and Google Ads - especially since they give coupons like candy.
In the end, it is still glorified advertising, while IMHO the core purpose of a business is revenue!
Advertising is very important, but securing paying clients does not depends nor requires advertising.
IMHO again - advertising however is a tool for growing, but just one tool among many others
This whole thing must have taken say half a day total - to say whether it was a good or a bad thing for them, you must consider the opportunity cost
And IMHO again, diverting just one mind away from important problems (to fix side issues instead) is a bad idea.
Focus matter the most.
In the end, it is still glorified advertising, while IMHO the core purpose of a business is revenue!
Advertising is very important, but securing paying clients does not depends nor requires advertising.
IMHO again - advertising however is a tool for growing, but just one tool among many others
This whole thing must have taken say half a day total - to say whether it was a good or a bad thing for them, you must consider the opportunity cost
And IMHO again, diverting just one mind away from important problems (to fix side issues instead) is a bad idea.
Focus matter the most.
New guys starting a startup that want to make their website "viral" look at this post, and learn. Give something of value, something that I can really learn from, maybe even not directly related to your startup, and people will give you credit. Not everyone can write a good post, but when you do be sure that I will check you website to understand what do you do. Also a couple of suggestions, your logo should take me to the website landing page, not blog home, this guys did a fine job by putting both links. And don't be ashamed of starting with something like "We at ZippilyBooppily(link) do.." this is perfectly fine, I know that you want my attention for your startup, and if you have something interesting to say I will check it, that link helps and tells me about you. Last and maybe most important, don't spam me, or HN or whatever, write only when you have something to say.
An article about gaming Hacker News is now the top post on Hacker News. I don't think that is a good sign.
He is not so much gaming HN but rather HN readers.
There's insufficient data here to show the new title resulted in the article score. It's a common trait on these kinds of sites, where an article's chance of "success" is massively improved after receiving only a very small (2-5) number of votes.
It's perfectly possible the original articles slipped down the new page too quickly before they got lucky and found a few interesting readers.
It's perfectly possible the original articles slipped down the new page too quickly before they got lucky and found a few interesting readers.
good to know HNers react to the same class of stimuli as the average google search user. I might try some Cialis-entitled posts
Here's 12 weird old haskell tricks discovered by a mom!
OBEY! What Hacker News doesn't want you to know!
Huh, an interesting concept. I wouldn't be surprised if, for certain products, HN hits are higher-quality hits than the average (higher conversion rate, possibly higher engagement rate after sign-up).
Of course, this is the sort of thing you need to be measuring. If that's not the case, you absolutely shouldn't be spending your money and, more importantly, your time on vanity upvotes and hits that don't result in a tangible benefit.
As an aside, I'd love to see a blog post from a generous individual willing to share how HN referrals as a cohort tend to perform compared to other traffic sources for their product/service.
Of course, this is the sort of thing you need to be measuring. If that's not the case, you absolutely shouldn't be spending your money and, more importantly, your time on vanity upvotes and hits that don't result in a tangible benefit.
As an aside, I'd love to see a blog post from a generous individual willing to share how HN referrals as a cohort tend to perform compared to other traffic sources for their product/service.
HN does not generally result in a lot of targeted visitors who are going to be long term users of your product.
However, I find the feedback in the comments section to be hugely valuable. HN readers are inevitably pretty savvy about startups. While you can get a lot of criticism - it can be quite constructive, pointing out things everyone else you show the site too has missed. Getting onto the front page is vital to get any volume of comments.
However, I find the feedback in the comments section to be hugely valuable. HN readers are inevitably pretty savvy about startups. While you can get a lot of criticism - it can be quite constructive, pointing out things everyone else you show the site too has missed. Getting onto the front page is vital to get any volume of comments.
I love this idea, very clever. A small spend can indicate a good option to you, and the traffic HN can generate is well worth it.
The spend isn't even a complete waste of money either, it's still pulling in a few new visitors during the test.
Very smart!
The spend isn't even a complete waste of money either, it's still pulling in a few new visitors during the test.
Very smart!
Thanks!
I was hoping it might be helpful to other people launching their startup. If you're already running Adwords, then that's a great place to start looking for guidance about how to market your product in a single sentence.
I was hoping it might be helpful to other people launching their startup. If you're already running Adwords, then that's a great place to start looking for guidance about how to market your product in a single sentence.
If hit count is your leading metric, then sure. But if you care what happens afterwards then you need to track what happens afterward, and see how the traffic clicks interacts; Quality over quantity..
My impression is that ranking algorithm, as described in the HN FAQ, is too simplistic, and as a result luck is too important factor in many HN submissions.
The more advanced algorithm would be for example to identify a set of HN users that actively up vote stories, and do not only follow links that already have a lot of votes; each new submission could be then injected to the front page of several random users from such set. With such approach, all new stories would have more or less equal exposure and equal chances of being noticed.
The more advanced algorithm would be for example to identify a set of HN users that actively up vote stories, and do not only follow links that already have a lot of votes; each new submission could be then injected to the front page of several random users from such set. With such approach, all new stories would have more or less equal exposure and equal chances of being noticed.
can't find the thread but i remember we sort of "agree" (lack of better word) to help upvote Show HNs at least to give credit to those who ship. i think if a submission (1) has Show HN (2) and really showing your newly launched product, it deserves an upvote
Editing flub?
'If you do wish to dispute this type of payment, we strongly lodging a ChargeBack with ChargeBack.cc so that we can optimise the speed and convenience of your refund'
https://www.chargeback.cc/paypal
'If you do wish to dispute this type of payment, we strongly lodging a ChargeBack with ChargeBack.cc so that we can optimise the speed and convenience of your refund'
https://www.chargeback.cc/paypal
If that is a screenshot of the experiment you were running, you just wasted $70. None of those differences is statistically better than the other one.
If you got benefit out of being #2 on HN, good for you, just say so.
If you got benefit out of being #2 on HN, good for you, just say so.
I think there’s real value in that it lets you think or other candidate titles which I would have originally dismissed as too simplistic or too boring. I'm not always going to pick the top one, but some degree of confidence and guidance about what resonates with people is better than nothing!
Also there were a lot of other titles with far lower results than are visible in those screenshots.
Also there were a lot of other titles with far lower results than are visible in those screenshots.
I believe he's saying you didn't gather enough data, which was my first impression.
I have also done this for years (not for titles on HN though -- good one), it's just that isn't enough data to really tell.
I have also done this for years (not for titles on HN though -- good one), it's just that isn't enough data to really tell.
I wonder if the other parameters were controlled in this experiment, like time and type of day of posting, amount of votes of other articles in front page, etc.
HN is already over-optimized for good headlines. Voting up after clicking is harder to implement but would improve quality.
You just invented A/B testing. Congratulations.
so meta
0.15% ctr is beyond shit
1. Write lots of blog posts on various topics. Write multi kilo-word reviews of honest to goodness books with reference to the ideas therein. Really pour yourself into the work -- take as your inspiration Michel de Montaigne's Essais. Submit some of these. Watch them get, at most, a single upvote.
2. Write one (1) angry rant about a company well-known to HNers, with no goals other than to vent a bit. Add "sucks" to the title somewhere. Watch as someone else submits it and your primal screeching zooms to the top of the front page and stays pegged to it for days.
Optional step 3: cultivate cynicism about HN and your reasons for sticking around.