Hacker News new | past | comments | ask | show | jobs | submit login
How We Tested Our Food Blog’s Expected CPM with Premium Images (cucumbertown.com)
64 points by Cherian on Aug 11, 2015 | hide | past | favorite | 32 comments



A little OT, but of all the Show HN's I remember, this one was not one I expected to see still going forward 3 years later: https://news.ycombinator.com/item?id=4669676

It looks like the pitch of being "Github for food" didn't pan out (or is de-emphasized), but I do like that besides being an attractive blogging platform, the site uses the recipe schema in a non-obtrusive way and has its own cooking-video-app. Great to see this still living on.


Thanks so much. Means the world. I quite don’t fancy the “fail fast” Silicon Valley model. Everything meaningful and beautiful takes time and sacrifice and lots of course correction. We also got acquired by Cookpad along the way [1]

We moved on from the Github or food to a Food blogging platform. Our users were using the platform to start writing and once they reach maturity we saw a migration to wordpress and other platforms. We knew we could fix that problem.

Also, by that time we realized we could do a much better job monetizing these blogs and making it into a livelihood by economics-of-scale.

[1] http://economictimes.indiatimes.com/small-biz/startups/japan...


This is the story of how DLC is breaking into the blogosphere. As a gamer this makes me want to run and cry.

But on the other side this is a more direct monetization scheme than advertising, as it gets 'round the advertiser middle man. As an internet user it gives me confidence.

I'm suddenly visualizing needing to pay $0.05 per image I want to upvote/downvote on Imgur.


Don’t take this literally. We don’t like the ad model much. Its very unobtrusive and compromising to the user experience. The authors spend a lot of time creating content. And they need to make a livelihood to sustain their passion. For that, as a platform we need to help them build such mechanics.

Internally, we debated putting a “refund button” right next to the pay button. We think that would create much more confidence. But hey, this is the first iteration.


Misleading title. They didn't double CPM, they ran an experiment that suggested they might be able to double CPM.

From experience, the number of people who click the "buy" button (even given the price up front) are not necessarily the same people who will actually follow through with the purchase. (Further, some people who are turned off by the "buy" word, are willing to follow through with a purchase when presented with a different initial choice; e.g. "unlock now")

It'd be more interesting to see the experiment run to completion, actually taking the money, and then see those numbers.


It could better be written as "Effective CPM". Also, the completion % is an assumption. If I take 10% completion rate, it is still higher than AdSense CPM rates and at par with other food blogging ad networks. For providing an ad free experience, it still seems worthwhile.


> If I take 10% completion rate, it is still higher than AdSense CPM rates and at par with other food blogging ad networks. For providing an ad free experience, it still seems worthwhile.

This is the much more interesting bit. It would still be super interesting to follow through and find out what the actual completion rate was.


As a food blogging platform trying to monetize and build a livelihood using author -> audience models, this is one of the first experiments that we are running. There are quite a few challenges and probably we fail a lot but this experiment seems like it could work.

Part of it was also inspired by Blendle’s experiment[1]

[1] https://medium.com/on-blendle/blendle-a-radical-experiment-w...


Let me suggest you run your experiment again. Use a different value amount, something higher than 20 cents. Consider actually charging the money via PayPal or Bitcoin or whatever. Yes, if you charge say $1 via PayPal you get something like 60 cents of that. If you get enough volume, so? And you need about a third as much buy-in to do just as well in theory as your 20 cents amount, a lot less to do even better in reality since all of those figures are phantom dollars that never actually got paid. If you are really committed to keeping it a very small amount, charge 20 cents or 25 cents per recipe but get your dollar upfront and give them credits for viewing other pics on other recipes. This might also incentivize return visitors. Some people will come back just because it is killing them to waste their remaining credits. Then you might hook some of them.

I read this with interest because I have several blogs. Some have ads. Some don't. Some have tip jars which used to be donate buttons. I have never made much money but I have always done better with getting cash from readers than with ads. I am starting the process of looking at Patreon as one possibility.

I am a woman and I started a food blog in June. So I really was excited to see the title of the article.

But having read it, I think your problem is basically mental bias. Recipe blogs are caught in an internet Pink Collar Ghetto. You are helping to keep them there by trying to find a way to charge a pittance and lamenting the lack of a small payment platform instead of putting together something actually viable in the here and now.

I served as a moderator at one time for a forum that was the biggest thing in its niche. The forum owner wanted it to make money but his mental models had a stranglehold on the income stream. Nothing was really acceptable to him. I see shades of that in the reasons you keep giving for why this, that or the other won't work.

Repeat your experiment with whatever modifications are necessary to make this real and not theoretical. Come up with a payment platform of some kind. Charge actual money this time. Improve on the model from there.

Best of luck.


Honestly it is stuff like this that Bitcoin would do amazing for if it had wider adoption. Scan the screen or click a button, send $0.10 your way, images appear. No muss, no fuss, no CC processors. The biggest benefit for this is, for someone who already is part of the Bitcoin ecosystem, the barrier to pay to see these images is smaller than using a credit card and maybe even smaller than PayPal. This biggest problem with using Bitcoin is that not many people (particularly those who would read a food blog) have it, and the barrier to buy into the Bitcoin ecosystem for many people is very high.


I'm not so sure. The real cost per transaction is astronomically high.


I'm not sure how you 'estimated' the 30% Payment Completion numbers, but that is way too high. That's 50% of your Payment Initiates. According to Wikipedia (https://en.wikipedia.org/wiki/Abandonment_rate), you should expect something more along the lines of 20%, not 50%. This puts your effective CPM at $3.10 rather than the $8.05 you claim.


I had never heard of cucumbertown before, so i did a 3 second lookover.

Quick question. It looks like you guys are trying to build a community of bloggers, who will probably be building a database of recipes.

If you're not doing it already... can you PLEASE have your bloggers add ingredients in a structured form. That would allow you to build an API later.

Imagine how much traffic you can drive to your customers sites if outside apps had an easy way to search through recipes.


We are not building a community of food bloggers. We are a food blogging platform. Think of us like Tumblr, specific to food blogging.

Some of the blogs on the platform: http://indugetscooking.com/ http://sousvidelife.com/ http://richagupta.cucumbertown.com/ etc.

And all the recipes are in hrecipe format [1][2]. And yes, we have an API that opens to cookbook companies.

[1] https://magazine.cucumbertown.com/taming-google-rich-snippet...

[2] https://developers.google.com/structured-data/testing-tool?u...


Why not say that, by making a $2 payment, you will unlock all recipe images on the site for 48 hours. With $2, you are still at a very low price, where it's disposable income for most... but offering a carrot that may bring them back to interact with the site more.

You might also try a $24/year level, or something that seems relatively inexpansive as a followup.

If you do this, you may want to combine with a social network login (google, yahoo, ms, fb, etc) with a minimal access (real name and email only). So they can access from other devices.

It would then be closer to a membership model, but the cost of entry would be very low.


My first thought on seeing this headline was: you're running your blog on CP/M???

Sigh. TLA namespace overload is the curse of our time.

(I would actually be interested in knowing if they make a go of this, because despite what I said above, the true curse of our time is advertising; anything to make it go away. I am concerned that their UI for encouraging people to pay looks rather sleazy, and would be inclined to drive me off, despite the trivial amount of money...)


I wonder how an incentive ad wall would perform. Users are much more likely to watch a video ad/download a mobile app than take out their wallet.


We are approaching the experiment with a philosophy of no ads. As the platform evolves, we could test incentive ad wall kind of solution on some other bench of recipes. Thanks for sharing the idea.


It sounds like they're going to have to go to a credits model. You buy $10 worth of credits and use those to unlock images.


We are trying to avoid "loading a wallet" concept. It adds whole lot of other complexities. This microtransaction is clean. No future committments. Now only if payment gateways made it viable.


You could try the reverse and float visitors credit. Once they're 5$ in the hole they can't view images anymore until they pay off the balance or something similar. Of course you'd have to find a way to make it so users can't just create a new account after they hit the 5$ limit (possibly by requiring a credit card and deduplicating by that?). Digital Ocean uses a similar model where they automatically bill your credit card once a month for accrued charges.


Are there any examples of that being used successfully? My guess would be that people will just create new accounts. This might be the perfect place to test that model, though. The content is already free, you don't have to worry about losing royalties that people don't pay for if the test fails. There might need to be a strong emphasis for having an account (being able to customize or save recipes without losing data from creating a new account).


I like this idea. You could present it as $5 of free recipes. It's like a try before you buy, or how casinos send people $100 free gambling credits. Yeah, some people will abuse it, but it's worth a try.


For the micro payments, I don't take the initiative to suggest bitcoin that often, but it could be a decent payment option if this were actually implemented.

Another social experiment you could try is offering a year subscription for something like $5, it might drive return readers aswell as solve the micropayment issue.


Great article. However... $('img').attr('style', '-webkit-filter: blur(0);')


Quick and dirty, you see! It's an experiment.


If a user realized that the "cost" of viewing the image was free, wouldn't the user disregard the cost on future instances of the experiment?

Did the experiment limit a user's participation to one time? Otherwise, the results might be skewed and the benefits exaggerated.


Yes, it limited the participation to one time. We don't show "locked" images, after doing unlocking. Most visitors are expected to be one-time visitors, finding the recipe via Google Search. Even if they re-visit, IMHO, it is measured risk to take in order to get actionable data for taking further decisions on shaping this feature.


I wonder if LevelUp could develop a service to facilitate these types of micro-transactions. I use their barcode-based app for "semi-micro" food purchases frequently.


Very cool. Would it be too hard to just take the users cc info, and then not charge them until their balance became a bigger amount?


Yes Josh, we could do that. That's a cool suggestion. In my observation, it is a low chance that a visitor will repeat the transaction on some other recipe. Most visitors come via Google Search and may not return ever again.


like many others, i too have never heard of cucumbertown, although i cook frequently based on downloaded recipes. i google for random recipes often, and not once have i come across this site. SEO and SEM might be something to think about, if this is truly world's only food blogging platform.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: