Home > Mobile > Follow-along Campaigns

---

Post-mortem, OLX Android IN (7)

---

I'm stopping my first mobile campaign. The offer, through Avazu, was for the Android install of a free app called OLX - the accompanying app for a free classifieds website. It ran it in India. The traffic source was Decisive. This is a direct-link campaign.

My total spend has been just under $32, which isn't a lot, but the stats are telling me that I can be 95% confident I'm losing 90% of my money or more, which I think warrants stopping here.

I was very surprised at the low number of clicks vs. impressions. I assume that's a reflection of me being not able to make a compelling ad. But I'm also surprised at the low number of conversions vs. clicks. Once someone clicks an ad that says, "Get the free app," and is taken to the Google Play page, I would expect a better than 0.09% chance of him actually installing.

My ads were 100% English, which may have been a problem. I'm considering running a campaign in Hindi just to see if it makes a difference, but the results are so dismal that I'm leaning towards dropping this offer altogether and moving onto another.

Digging through the segmentation reports, I only found a few interesting things: 1) People in Punjab seem to click more often than anywhere else (CTR ~4%), 2) those on a site called pokemoron.com click a lot (14%), 3) those playing ViMap games click a lot (~3%).

As a whole, my results didn't seem to vary much between app/wifi, app/mobile, site/wifi, site/mobile. All four were equally shitty.

I'm also concerned about the reliability of my tracker, Voluum, because it looks like I'm losing people on the redirections to and from Voluum. The clicks reported by Voluum are smaller than those reported by Decisive. And clicks reported by Avazu are smaller still. http://bit.ly/1pWjP5b

Before moving to another offer, I'd like to address the tracking problem. I'd prefer to move onto another offer with more promise before testing how Hindi or a landing page affects CTR and conversion rate. But if anyone has any other advice, I'd love to hear it. Thanks and good luck to you in your campaigns too.

---

I've never used Decisive as my traffic source, but in my experience India is a very tough place to make money with any offers. I've made money elsewhere. Not in India. I would stay away from India if you're new to mobile camps. I'm sure there are ways to make money with offers in India though.

---

I'm not so sure about this. Why would running an offer in India be any more difficult than running the same offer in another country? If you do a little bit of research on India's culture, economy, mobile market, and it's people you should be able to figure out at least a few killer angles for any offer. At the end of the day Indian mobile users have needs just like us...


I’m still learning and I’m far from being an expert but from what you’ve posted so far it looks like you’ve done 1% of the work before wrongly concluding that the offer and/or geo is a dud. There’s a reason why every mobile heavy CPA network has this offer - probably because it works…. If other people are making it work then why can’t you?

Firstly, did you do any research on the Indian mobile market and its users? Here are a few quick facts:

Mobile phone ownership and usage:
- The greatest number of respondents, 83%, came from the age group of 18-34
- Among those surveyed, 77% males dominated the viewership, while women formed 23% of the viewers.

Source: http://yourstory.com/2014/04/mobile-video-survey

From the above you can reasonable conclude that that India’s mobile market is HEAVILY dominated by younger people. I’ve seen lot of other data/studies that confirm this. Keep this in mind as we go to the next data set.


Demographic Info
- By age demographics, 54 percent of India's population is under the age of 25 - that’s a whopping 550 million young people.
- India is certainly the world's second largest English speaking country after the USA (300 million - near 100 percent of US population).
- De facto, English is the "common" language among most educated Indians today. When two Indians from different states meet, they invariably communicate with each other in English.
- English is the language of all businesses, government and education.
- The entire university education in India is in English — very similar to what you find in the USA.
- Every computer user in India uses Microsoft Windows — English-edition. Microsoft has provided support for Indian languages only from Windows Vista released in 2006 and not even 1 percent of all computer users in India use Indian languages for computers. All emailing within India is in English.
- In every major city of India, the cable TV has at least 25 English channels. There are many Indian English channels and others are the American/Western channels.
- The Times of India, an English language daily newspaper in India, is officially the world's largest selling English language daily newspaper.
- Among young Indians, the percent of English literacy is much higher. English skills are a must to get an entry into corporate India.
- In India, English skills are part of the social hierarchy. Practically every person in the higher and upper middle class section of the Indian society is fluent in English. In fact, English fluency is kind of a pre-requisite for the upward mobility in the class ladder. An upwardly mobile successful Indian entrepreneur with inadequate English skills is not considered blue-blooded enough and is persona non-grata at the turf clubs or country clubs of India. A classic example is the detente between the power-flexing non-English speaking Indian politicians and the English-speaking movers and shakers from Indian bureaucracy and the industrial clans.
- The predominant role of English language in Indian society is making new generation of Indians to have near-native English skills — while the young ones are still in the elementary schools. A cursory visit to any of India's KG or nursery schools in the metro cities illustrates this fact.

Source: http://www.indiatribune.com/index.ph...try&Itemid=471


What we know from above is that the vast majority of mobile users are younger (under 40ish), and now we also know that most of the younger Indian population is heavily exposed to the English language on a daily basis. Do you still think that your English ads are a problem?

I’ll tell you right now that there's definitely nothing wrong with Voluum or the offer (in India or other GEOs), and Avazu's conversion rate for OLX is generally equal to or better than other networks.

Also if you're angle is something similar to what the stock creatives are using then that's your problem right there. I’m not running OLX anymore so I’ll share a little bit of what I have to help you out. Here are a few notes I took before running the offer:

Advertiser Headlines/Angles
- What are you waiting for?
- Free classifieds on OLX.
- Buy what you’re looking for
- sell everything in OLX
- India’s largest marketplace
- free classifieds
- sell your old xxxx (they used items like bike, tv, etc.)

Advertiser CTA
- sell now
- get free app
- publish free
- download app

Generally speaking, their angles advertise “free classifieds” without actually giving a good reason to download the app. While I was brainstorming I put myself in the shoes of someone who might see one of these ads and I thought to myself, “would any of these angles convince me to download the app?” The answer was an obvious “no.” This is an app to buy or sell goods so think about items that might be popular with the younger Indian population. Think cell phones, tablets, etc. On a higher level, think of an angle that will change their life for the better and will actually have a positive impact. In highly developed countries we take having an iPhone or Samsung Galaxy for granted, but in countries like India it's a big deal to own those kind of phones. If you can help someone find a used phone that they otherwise wouldn't be able to afford then maybe they'll download your app.....

Here are some angles that I came up with. I’m not going to tell you which ones I used but all of the ones I submitted to the advertiser were approved.

My Angles - targeting buyers
- Samsung Galaxy - Cheaper price than store?
- Samsung Galaxy - Never pay full price
- Samsung Galaxy - Buy used. Save 50% or more?
- Samsung Galaxy - Buy with big savings
- iPad, iPhone - look for big discount
- Buy iPad, iPhone, and other electronics with big discount!
- Samsung Galaxy - Are you paying full price? No one else is
- Samsung Galaxy - Why Pay Full Price?
- Samsung Galaxy - Buy used pay no sales tax
- Samsung Galaxy - Find With Big Discount
- Samsung Galaxy - Buy With No Import Tax
- Samsung Galaxy - Buy With No Sales Tax
- Samsung Galaxy - Cheaper cost than United States?

My Angles - targeting sellers
- sell your old samsung - buy new one!
- upgrade your phone - sell your old one for big $$$

It’s easier to come up with angles targeting buyers, but for good quality you’d probably want to target sellers as well.

In regards to your data, if you break it down further, what do your numbers look like? $32 isn’t much for testing in India. Also where are those 1100 clicks coming from? If you have a few sites or apps sending most of the clicks (which I’m guessing is the case) then how do you know that the campaign or offer is a dud? Remember that you’re supposed to test each placement/app for at least 2-3x payout otherwise how can you make any statistically significant decisions?

Lastly, I really think you really need to use a landing page for these types of offers. I’m sure some can kill it with just banners, but from what I’ve found using a landing page helps a LOT. Put in the work and make a few QUALITY landing pages and I’m almost certain you’ll see more promising numbers right off the bat.

I feel like it's so cliché to say this but ultimately you should stick to one offer that you know is doing well, or a type of offer, and try EVERYTHING to make it work. Test different ad formats, countries, ad networks, ad copy, LP vs direct, bidding strategies, etc.

---

@zeroonedigital, thanks so much for taking the time to write this.

This is a great point. The hardest thing for me, as a beginner, is doubting whether an offer has the potential to be successful. Without confidence in the offer, it's hard to push through. It's like climbing without being able to see the top. But this gives me the confidence I need to commit. Thanks.

True, my best guess is that English would not be a problem in India. Being able to use English is part of the reason I chose to run the offer in India. But I think it's still worth split-testing at some point. For now, I'll take your hint and focus on the more likely problems.

You're right. Most of my creatives played the generic angle, similar to the stock creatives. One of my creatives did consistently better than the others. It had a more specific message: "Cash for junk". I had planned to test each vertical individually (mobiles, furniture, clothing, services, etc) if the first set showed promise. But I guess I cannot expect just any angle to show promise - I have to find the right angle.

Site traffic outperformed app traffic with an ad CTR of 0.85% vs. 0.49%. About a quarter of my site impressions came from a single source. Within the app category, the largest source accounted for 18%. However, in no configuration did the source with the highest impressions get the most clicks. I suppose this means if it continues that way to a point of statistical significance, then I should blacklist the sources that aren't performing.

The payout is only $0.58, so in total I've surpassed 50x the payout. But I didn't realize there was such a focus on matching results to specific placements at the beginning. I'll start examining my stats at that level.

My feeling here was to see how it does without a landing page first, and then add a landing page later to see the difference. With an ad CTR under 1%, I feel like I should be focusing on that first. But I'll take your advice here and create a landing page (or a few) to sell the install.

Thanks again for the advice. I really appreciate it. I'll get to work

---

@mikesf - Good thinking, and a very good idea to do a postmortem.

One thing to be VERY careful with on mobile campaigns: statistical significance calculators should be used with caution. The reason for that is that you're not actually running a single campaign: you're running a massive collection of campaigns across different placements, handsets, carriers, etc, and any one of those combinations could be successful.

So will you be successful running RON on India without pausing anything? Probably not. But will you be successful with one carrier, or one placement? Maybe.

That's why we give mobile more time and more money to see if it has potential - it's not as simple as running on POF, say, where you don't have any confounding variables making the entire thing confusing, and where your campaigns tend to be narrow by default rather than starting broad and narrowing down.

---

Thanks. I didn't realize that, but it makes a lot of sense.

---

Wow zeroonedigital !!!

You CAME in there like thunder mate, i sucked every single word on that post. Now i see why i failed with that
offer. Thanks again for that

---

Home > Mobile > Follow-along Campaigns