Hacker News new | ask | show | jobs
20% revenue split to help monetize a website
11 points by tomkazarian 4150 days ago
I have a site getting significant traffic and need someone who is knowledgeable about ad networks/ad layout testing/etc. I will pay 20% of whatever revenue your suggestions bring in for a period of 6 months.
3 comments

When it comes to ad sizes, you'll want to run 300x250 on your site using AdSense. It has the highest stock available, so it'll generate the most bids and profit.

You might want to consider randomizing the ad locations in your results. I viewed a few pages, and I already know where the eBay ads are listed, so I'm skipping over them. You could randomize an eBay (if they pay well) or 300x250 AdSense advertisement in every other row.

The best option is always selling the ads directly on your own or through an ad manager that takes a large commission. For example, I earn 20x or more on my direct ads, compared to AdSense. Of course, I don't always have a direct advertising campaign, so AdSense runs as backup when one isn't available.

Browse Commission Junction and try some offers, although I have low expectations. I usually send them a few hundred clicks for every sale, which only earns a few dollars.

Your FAQ says people are interesting in selling items on your site, but that's not available, you just scrap sales from forums. Why not allow people to promote their forum sales on your site? For example, on the top of every page, you have a featured row with 4 random sales in that category, currently being promoted. So, I'm selling a watch on WatchUSeek, I visit your site, pay $20 to get listed in the featured row on the watches category for a week. For example, here's a quick Photoshop...

http://i.imgur.com/uRBXAdD.jpg

Good luck, advertising is an uphill battle, and you need to spend a lot of time trying out different strategies and carefully monitoring the results.

Holy smokes- GREAT reply and complete with a mockup!

I'll start loading up some of those 300x250 units. I guess my hesitation in the past was that I didn't know if those ads would adapt to screen sizes/orientation and the units would break the layout but I might as well find out now. I'm still scratching my head on the gun section though since it's very poplar but I can't run AdSense in that category (against TOS), but I guess direct deals, CJ, and featured guns would work there. You mention an Ad Manager for getting direct deals... Is this someone I hire from a job board or are there ones you can recommend?

>So, I'm selling a watch on WatchUSeek, I visit your site, pay $20 to get listed in the featured row on the watches category for a week.

>I'm still scratching my head on the gun section though since it's very poplar but I can't run AdSense in that category

Popular gun forums are on the same boat and that's why their side-bars are filled with advertisers/affiliates/sponsors. Just contact those with your numbers on the sub-section, i'm sure you'll sell out quickly.

Your website has a lot of potential for targeted advertising.

Yup, you're right. Featured + advertisers found on gun forums should solve it.

Time for me to think/ask an advertiser about what he wants (i.e. ability to login and see impression/click counts, etc.) and what the CPM rate(s) should be.

I appreciate the time you took to think through that.

>Time for me to think/ask an advertiser about what he wants (i.e. ability to login and see impression/click counts, etc.) and what the CPM rate(s) should be.

The problem is often that we can't guarantee the stats like google does. (for clicks) That's why i either sell a guaranteed minimum impressions per month or a flat CPM rate.

Still track the impressions and clicks, but only as a indicator or email them reports to save the hassle of managing credentials.

The good news is that you're dealing with high priced items, so you can charge higher rates. Your number one selling point is going to be whether or not being listed in your featured section offers a better chance of generating a sale.

You'll want to track clicks for featured vs non-featured items. Also, you'll want to follow-up by visiting the forums for your featured items, and then compare those to the average non-featured item. Are the items you're featuring ending in a sale more often, or faster? If someone is selling a $50k car, and you can prove they have a 20% higher chance of selling by featuring on your site, then paying $50 to get featured is a deal. However, I don't know if you have the current traffic numbers to have that sort of benefit.

You can try the above before you have a single paying customer. Choose 4 products at random, feature them, watch the results and increase in clicks.

When it comes to AdSense, don't expect to get rich. I'd say your site would pull in $3-10/day.

For an ad manager, I do a 50/50 split for direct advertising campaigns. The number is high because the individual managing my advertisements previously ran one of the largest companies in my industry, and they regularly meet with the large billion dollar companies that I would struggle to get time with. An equivalent for you would be someone that has great connections at Ford, GM, Chrysler, etc, so when Ford releases a new car and is running an advertising campaign, the related category on your site is a part of that advertising deal. How you find that person, I don't know. In my case, they contacted me because of my user base. Figure out the CPM rates you can earn on your own, then you can pay someone commission for anything higher. For example, you might earn $0.25CPM through ad networks, and then you can sell direct advertising or $2.00CPM, or find someone to sell it for you.

It's a really tough business, I can't stress this enough. AdSense is the best ad network for general fill, so give them a test first. Try the affiliate offers through CJ as well, they might work better in your industry than mine. For a while I was earning $20k a year through CJ on the side, but it went downhill in my business.

You might also want to try accepting posts on your site. Allow users to upload a photo, title, description, and contact information, then mix those with the results you're scraping. Over time, you could possibly phase out the scraped content and focus on user submitted content, and adding additional categories. Then you'll have a little more control over the content, users, and charging fees. A site with original content is going to be more valuable than a site aggregating content from other sources. However, aggregating content is a smart method of starting such a site, so you might be in a good place to make that transition. Definitely consider this option, it might help your site grow into something much larger.

Edit: One last point. Visit your competitors, or other classified sites. They've been down this road before, so see what ad networks they're using, what ad sizes and placements, what offers they're pushing, and how they're charging users, etc. They've likely spent years trying to find the perfect solution for them, so use their knowledge as a starting point.

Thanks and great points!
Advertise on porn networks. Traffic Junky is dirt cheap compared to most other display networks, and these sort of things would likely appeal to the predominantly young male audience.

Another suggestion is reddit ads. Align the categories with various subreddits and write some decent funny copy.

Had some earlier success with reddit ads and even made it to the Top 20 (all time) on a decently sized subreddit that I manually submitted to. Never heard of Traffic Junky but it makes total sense. Thank you for the suggestions! Brilliant.
No problem at all. If you need any more help feel free to shoot me an email. I do lots of marketing for tech companies, and I'm always happy to give out some advice for free.
If you post the site url, I'm sure many of us would be more than willing to post some monetization-advice here for free. Just a thought.
Thanks for the suggestion.

---

At the moment, I'm running only ebay ads incorporated into the following areas:

- 2 units per page on category pages like: http://www.lionseek.com/watches

- 2 units per page on brand pages like: http://www.lionseek.com/watches/brand/seiko

- 12 units per page on individual item pages (found in the 'similar items on the web' section/slider) like: http://www.lionseek.com/watches/brand/seiko/fs-seiko-6309-50...

- 4 units on a search result page that has no matches like: http://www.lionseek.com/watches?seller_id=-1&has_photo=true&...

---

I have an Adsense account, Commission Junction account, and an Amazon Associates account it's just that none of these have been implemented. I can get self-serve ads implemented quickly. The site is responsive so the ads would have to work with a couple of breakpoints.

An ideal scenario is someone saying 'Stick a XYZ ad unit here stylized in the following way. Stick ABC ad unit here in the following format. This ad platform (XYZ) might help for testing and maximizing revenue.'

If you are confused about what the site is/does then take a look at: http://www.lionseek.com/p/faq

Any suggestions would be TRULY appreciated!

In your shoes, I think I'd start by incorporating generic adsense units in order to get a "baseline". It'd probably be easiest to start with the newer "responsive" ad units that Adsense offers, since those will fit the container you make available, and you won't have to worry about breakpoints and such.

I would run those plain ads for long enough to get statistically significant data on CTR and CPC. Adsense revenue will likely vary widely for a few days and then level-out in a consistent range after a while.

Once you've got that baseline, you have a metric by which you can compare other strategies and options. Then integrating something like CJ/affiliate ads becomes simpler. You just let that new strategy run for a small subset of visitors and see how the revenue compares to your "control group" (the visitors who saw adsense). Then you can be methodical and scientific in your approach, and test all options you have available to you. This is the same method by which you can test other adsense units, other styling options, etc. Split-testing software like Optimizely makes these kind of tests very easy to run and get data on.

Good suggestions - thanks!
The site seems to be marketing products that are targeting men. Perhaps the it's design can better reflect that? Maybe it can better retain customers. Right now it looks a little 'generic'. I noticed the colour scheme for the search page seems inspired by google. I've made <$500 in adsense revenue so no expert here.
You're right - the categories currently on the site are usually more popular with males than females.

To be honest, the design is the result of me fumbling around photoshop. In the future I hope to get a real designer involved ;)