Affle India - India Mobile Internet Advertising Leader

Few things stood out for me in the call -

  1. Contribution of acquired entities in the H1 PAT was hardly 8% with PAT margin of 10% so their organic business is rocking and has much higher margin. This presents an opportunity to turn around acquired entities.

  2. Data privacy - They are well prepared and waiting for more stringent laws as they already comply with Singapore laws which was the among the initial countries to bring data privacy laws. Anuj sounded as if this would benefit them and hurt some other non compliant ones.

  3. Growth - H2 will be better compared with H1 due to seasonal factors. He also indicated that emerging market (ex-India) could growth 25% CAGR for the next few years. India might grow faster. They also feel growth India could pick up further if economy grows faster.

  4. Financials - Working cap to improve substantially in H2 as they start collecting. They aim to finish the year with 70-80% of all collection by the end of FY.

  5. Costs: No big employee growth as such but acquisitions will bring more employees. They keep looking for more.

Disc: Invested

6 Likes

image

12 Likes

Sorry! Didn’t understand the logic. Which stock remains at IPO price after listing? It will go up or down based on change in fundamentals or just perception. BTW,Anuj did mention in one of his pre-ipo interviews that bankers asked him to tone down his valuation expectation due to market condition. This Twitter statement reflects poor understanding of investment banking. Valuation at the time of IPO and sell side research is not done by the same person due to Chinese walls etc. even if done by the same analyst IPO price is also a call taken by the management.

5 Likes

This could be biased statement …
I think the twitter acount is perfectly fair…
what great things happens with the company in 4 months so valuations get doubled…
It happens only in india

2 Likes

I have been reading about Affle (DRHP, Nomura Report etc.) and have following questions about the business.

The first thing I am most interested in is Affle’s relationship with Google/Facebook - as publisher & as competitor & as OS provider for the device.

With Google, Facebook as publisher - only 5-10% of the ads are provided by third party tech platforms & rest are provided by in-house ad tech teams of these giants. How does one see this space evolving?
Can share of third party vendors go from 5-10% to 20%? Why would this happen?

The Nomura report claims that time spent on Google/Facebook is likely to reduce in favor of third party apps. This is the trend seen in the US. This trend needs to be tracked.

With these giants as competitors, why would a publisher or advertiser go to Affle vs. say Google or Facebook? May be Affle is small and can provide more customized services? But quality of data/insights might be far better for for Google for instance. Why choose Affle over Google or Facebook remains an important question for me.

To take this competition aspect forward, does someone know if Google/Facebook work on performance based model (CPCU) or they still work on impression based models (CPM). Nomura report claims that only 10-15% of the ad tech market in India is performance based & Affle model is likely to force everyone else to performance based model. Is performance (CPCU) based model really an competitive advantage for Affle & for how long?

Bringing in further points, how do CPCU based pricing gets dictated in the industry? From Shemaroo experience, it looked like CPM rates moved down for YouTube over the years. What happens to CPCU rates if Google/Facebook reduce the rates?
Can Google ask for a cut from players like Affle? Can Google hinder Affle from collecting data?

The second thing that hit me from Nomura report is that Affle might have minimal presence in Video ads. Does Affle have capability to place video ads on platform like Times Group? How about YouTube? Given that video is one of their strategic points & there is gold rush in video viewing, it is important to get split of video vs. non-video revenue.

Finally, the question of Amazon - Amazon has 7% market share & it is third largest ad tech player in the US. Amazon has launched ad tech services in India. What happens to Affle over the long term?
The broader question is - Why would any publisher with large amount of data (Hotstar, Jio, Flipkart, Times Group etc.) will not create its own ad tech platform? Currently publishers seem to get 40% of the ad spend in value chain. Why not take that number to 70-80%?

Another aspect is wide variance in monetization factor in India (6.9%) vs. EM (1.8%) or DM (0.9%). Most people believe that monetization factor in EM or DM will improve to India. My question is - can reverse happen - where India monetization factor goes down? Why is it high in India? Is it simply matter of focus?

I am also interested in numbers & capabilities of InMobi & other players - if someone has done work, kindly do share.

Disc - No investments, not a buy or sell recommendation.

14 Likes

Having seen Digital Marketing life cycle and spends briefly - there are below key players broadly speaking

  1. Agencies - content, creative, campaign etc.- more of services characterized
  2. Publishers, Exchanges etc - ad inventories etc - more of sell side - Times group, Google and FBs of world( one facet as their biz is humongous)
  3. platforms with Ad Tech/Mar tech - Again google And FB as well as Affles of the world.

Now, we all can observe that money flows from brands to above 3 in some mix and configuration. In india 2010 onwards there is decent mktg budget growth with digital pie growing at higher pace( noteworthy that mobile initially was smaller share in digital but should be lion share as we go forward).

Coming to Affle, positioning aspects they got two solid tailwinds -firts thing is to be specialist in Mobile focused advertising( smart move, though they may have started as generalist), with timing in their favor( mobile usage boom, cheap data and thus low spend cost) and an outcome based pitch of CPCU which sounds disruptive( most Agencies preferred clicks/impressions as well as omni channel campaigns).

Technicalities aside, Brands and Agencies love outcome based pricing and Affle just did that at a scale and profiling engine and ML/algo claims( still to get convinced here).

Lion share of Affle comes from Agencies and per latest concall tech savvy brands increasingly turning their direct customers.

This is a very fragile space - as in if campaigns dont deliver - Agencies or Ad tech partner will not get enough spend allocation- so must give that Affle doing well and delivering to clients must be happening effectively else numbers wont simply come the way they have.

For Short to low mid term - There doesn’t seem a comparable competition to Affle, aided by network effect of higher usage delivering more profiles which in turn helping outcome quality for clients and non linear co relation with expenses( platform aspects), they seem to be only sizable contender with scale.( lack of research here on their profile engine)

Coming to google and facebook, they have larger challenges/opportunities to focus - Affle in no way seems a threat - infact it could be an acquisition target should they want to be closer to agencies and brands.

For sure being a Tech platform, it is bound to have disruptions( content moving to Video, subscription based usage, VR and AR catching up…), this is where it is imp to nibble relevant skills as Affle grows- and they seem to be on path.

Party is there to last for some time given mobile usage growth ( few quarters or longer is anyone’s guess) given tail winds but dont see this as secular story yet.

One key thing to observe would be any sizable uptick in floor prices for telco leading to higher data prices may play spoiler as it may affect mobile data consumption growth to some extent( hence price dynamics which is skewed to digital mobile ad performance). Probably a risk for entire sector.

Discl: Tracking position and interested.

14 Likes

As per BSE, Affle wish to inform that the following Analyst/ lnvestor Meetings of the Company have been scheduled, upon request received from the below entities.
investor_opportunity.pdf (133.4 KB).
Is there any opportunity here?

This has been a very useful discussion so far, there was a recent presentation by Vertoz at the Alpha Ideas SME meet. The company representative gave a good overview of the adtech industry. You can find the video at the link below:

1 Like

The menace of Ad Fraud in India

According to Moneka Khurana, country head for the Mobile Marketing Association in India, ad fraud is a huge problem for the entire digital advertising ecosystem but especially in India where ad fraud has risen by as much as 40%.

The MMA conducted a survey to benchmark the current practices that marketers use in order to assess and combat mobile ad fraud.

Highlights from the survey include:

  • Although 22% of mobile ad budgets fall victim to ad fraud, exploring new technologies or approaches to counter it is not a high priority for most marketers.
  • Nine out of ten marketers are hiring dedicated external vendors, or are planning to do so in the near future, to detect and prevent fraud.
  • The top three areas in which ad fraud is highest are: brand campaigns, social and publisher direct.
  • Most marketers expect real-time analysis and proactive blocking of fraudulent impressions. However, service providers often offer periodic analysis which seems to be less important to marketers.
  • Marketers are not very familiar with blockchain, but almost all of them believe that it has the potential to help against ad fraud.

According to another study conducted by techARC, the technology analysis, research and consulting firm, at $1.63bn, India’s share in global digital ad fraud stood at 8.7% in 2018. It is expected to grow to 23% in 2019 as a result of sophisticated techniques being employed by fraudsters. Digital commerce contributed 51% of the total ad fraud in India.

Although app fraud contributes to over 85% of total digital ad fraud, web platforms are more susceptible to frauds because in several organisations the digital teams are primarily focusing on the app, leaving the webspace vulnerable. There are also several brand safety issues specifically emanating from the web, such as fake leads and keyword abuse.

Lastly, as video is becoming the preferred medium of content in India, it is attracting fraudsters to explore sophisticated fraud techniques to earn more on the premium advertising channels.

The Mobile Marketing Association (MMA) India’s ‘Ad Fraud roadshow’ brought together leading marketers and subject experts in order to spread awareness and to highlight the several key challenges related to brand safety and accountability in mobile advertising, and to explore solutions for ad fraud. These roadshows addressed growing concerns in the mobile marketing industry and it was generally agreed that building standards, guidelines and best practices is the need of the hour to deter ad fraud.

Uday Sodhi, digital business head at Sony Pictures Networks India, said that he does not believe too many CMOs pay attention to ad fraud and the amount of leakage that it can cause: “I would be very worried about ad fraud [given] there is lack of awareness at the clients’ end… when they say that they have put in INR100 but actually only INR50 is going into a valuable campaign and INR50 is really getting leaked out - that leakage is what the client is not aware of,” said.

“I think there is enough technology today in terms of protecting yourself but we are ignoring ad fraud when it happens. Increasing the security of every transaction is a learning process for everybody – whether it’s the client or the publisher or the intermediaries or the agencies.”

According to Jahid Ahmed, head of digital & content marketing at India’s HDFC Bank, the ad fraud situation is far more grim. He said that the overall digital ad spend in India is $2bn but of this a staggering $1.6bn is wasted on fraud.

“Today, bots can imitate to an extent where they can actually talk to a call center as well, and so money is being wasted and the call center is wasting its bandwidth,” he said.

Five ways India is battling the problem

  1. SUPPRESSION Ahmed said that at HDFC Bank, 80% of their ad spend in India is on digital and 80% of that spend is vi mobile platforms. Understanding and combatting fraud is crucial.“We don’t believe in aggression in terms of leads, what we believe in is campaign orchestration where suppression is a key element. For example, on a shopping site, if you notice that even after the shopping is completed you continuously see the ad again and again, you get irritated. Suppression is important. I also think what is required is a unified, seamless measurement of the unique result or reach on investment.”

  2. ALL-ROUND EDUCATION Deepak Tahiliani, head of digital trading at GroupM India, the advertising media company, said they take a holistic approach to solving the problem.“First, it’s about internal education. We ourselves need to learn about what we need to do – as an agency we are custodians of the brands we handle, so we first need to step up our game. Then we need to work in parallel with clients to educate them on how this needs to be controlled to boost their confidence. We can follow a robust mechanism to educate clients on how we use third party tools to measure and then control fraud,” he said.“On the media vendor side, we are educating them to step up their game. Sophisticated fraud is rapidly changing every day, it’s not a one-off activity. It has to be a constant build-up of education, awareness and implementation across the ecosystem.”

  3. SHARING DATA Vipul Kedia, chief data and platforms officer at Affle, the mobile advertising company, believes the most effective way to control ad fraud is data transparency.“I know that any brand or organisation would want to hold their data close and would not want to share it. But the best way to combat ad fraud is to allow people to measure their own matrices, wherever the publisher sources are. If we know what is the kind of efficacy we are driving for the brand, then I would automatically know where the spends are being effective or not being effective.”

  4. DETERENCE IS KEY At the Bangalore chapter of the roadshow, Venkatesh Kamath, associate director at Flipkart, had stated that it would be too optimistic to say that this problem can be eradicated.“I don’t think the solution is elimination, I think the solution is deterrence. We make our deterrence policies very clear, we have taken a lot of steps which have ruffled a lot of feathers, but we have been persistent. But we want things to keep evolving, we want data to speak for itself, we view a whole lot of data to make sure we are able to detect fraud.”

  5. A NEUTRAL UMPIRE Dhiraj Gupta, CTO & founder of mFilterIt, the fraud detection and prevention company, took a contrarian view and said that publishers and agencies or anyone who is getting a part of the ad spend pie cannot be expected to take responsibility of this problem because they have a vested interest in the issue.“Platforms will be more interested in ensuring that they have a higher media share in the ad spend, so I think a neutral, independent validator or enforcer is critical,” he said.“Advertisers have still not matured enough to start asking the right questions, ad fraud is not just about bad traffic, it now also means organic traffic. I can have sales, leads and conversions that are fraudulent, what that means is fraud has moved to a level where someone can steal your organic traffic and show it as if they have done that advertising traffic for you.”

5 Likes

Hi All,

Disclosure : I have been part of the Mobile AdTech industry for 7 years before recently starting up on the SaaS space. I have worked with partners like Affle, Vertoz etc. and with the larger adtech giants.

While Affle is a profitable entity and showing growth, at current valuations I have a few concerns and would like to highlight the same for a better discussion.

  1. Affle’s significant revenues are through affiliate/third party network inventory instead of direct SDK/Owned & Operated properties or DSP with header bidding. In this form of inventory trading, they have little control and no visibility into the end user. Postback of deviceID/GAID adds almost no significance to user profiling. Further the tech stack for this inventory is a third party SaaS player and not owned/built internally. This also doesnt help when advertisers/FTC start looking into data privacy and if there is an explicit consent from user before ads are shown to them.

  2. The DSP business is only 8% of revenues, and as advertisers move to seek more transparency in inventory and Ad-Fraud reduction , ad tech players who have a better bidder, ability to inject organic data through DMPs will have better survival chances. eg : Google DBM, IronSrc, AppLovin, InMobi, TradeDesk etc.

  3. Why Outcome based pricing is not a moat or differentiator ? All AdTech players provide for some form of outcome based pricing. Including Google, FB, etc. We call this is CPX/CPA, while bidding is on CPM/CPC basis. But the bidder optimizes to convert to a CPX range. The reason why, players like Affle and 100’s of other affiliates like YeahMobi, Mobvista, SVG, Mobave etc offer CPX only instead of even CPI ( install and app open), is advertisers find them to be in the lowest rung on quality/fraud and do not want to pay until they are convinced the installs and end actions delivered are real. Further, through independent and Mobile Tracking/ Attribution partner based anti-fraud systems they are also increasingly called out for organic sniping/click spamming where the advertiser claws back payment for installs which might have been sniped/stolen from organic traffic or attributable to other channels. Further, mobile tracking attribution platforms used by advertisers like Branch.io and Appsflyer have a better understanding of fraud than independent players like mFilterate as they have a tracker SDK and way more data flow, buts thats a separate discussion. You can also check Appsflyer industry report by geography. Appsflyer is the largest tracking platform for mobile apps in India/SEA and is an Israel firm. Some of the current rules set by these partners allow the advertiser to retrospectively clawback for installs/actions delivered. Hence comes the impression transparency part, as unless you are in the direct impression path or own the inventory, there is very less visibility on the end user actions and higher likelihood of taking credit for organic users and being called out for Fraud. Further, advertisers in the past have taken legal route as well. https://adexchanger.com/mobile/is-ubers-new-ad-fraud-lawsuit-futile-or-game-changing/. https://www.appsflyer.com/resources/ecommerce/know-enemy-complete-mobile-ad-fraud-glossary/

  4. Which are the key channels are share of third party networks and who are they ?Google & FB dominate 70% of spends for large companies, for smaller ones its almost 95%. The rest of the spends is distributed amongst other walled gardens like Snap, Apple Search, Tik Tok, Hotstar, OTT players, Apps like Daily Hunt etc, Times Internet ( 10-15%) , OEMs ( Samsung, Xiaomi etc) and then comes the DSPs and the third party networks like Affle etc. With more access to direct inventory or exchanges which can provide transparent inventory coming in from mediation partners and lower end apps ( MoPub , iDSP, TradeDesk, AdX) and pushed by mobile attribution players to avoid organic sniping , third party affiliate networks would be more marginalized.

  5. If we see mature markets like US, affiliates have very low share of marketing dollars and in gaming industry dominated by players like Applovin, Unity, IronSrc and Vungle virtually there is no presence of affiliate/third party networks like Affle. These players have their own medition stack, ad serving SDK and they dont bring in any advertiser working with affiliates as there is organic sniping through creatives/ click firing. But then, why are they doing well in India/Asia or thriving ? My view is some digital marketers running the user acquisition budgets are naive or are clearly incentivized to show that there digital/mobile marketing efforts resulted in huge number of app downloads/actions irrespective of whether the installs are questionable or organic sniped. Also VCs push to show user growth and traction. Slowly but surely they will wake up and taps can go off, and further the the larger channel partners like O&O properties, Google/FB will also push back on their installs being sniped and push back on advertisers using these affiliates.

  6. Advertiser concentration risk is a worry. As of FY19, 110 cr of 250 cr revenue is from 2 partners and mostly the same advertiser ( Amazon). If Amazon like Uber or other large advertisers feel most Indians who would want to use their app is already an user, they would not spend further. Also, given the tech stack is external SaaS, few employees can move out and take away with them their advertisers and the supply sources. If the supply is exchanges like MoPub, there is practically no moat apart from working capital to access these supply sources through any other DSP. Same if the supply is other smaller affiliates or chinese partners like YeahMobi , Baidu etc which can be accessed with ready to use platforms like Affise, HasOffers etc… When Tyroo ( a similar network like Affle) sold to Dentsu several former employees started their own firms and are doing well (https://www.linkedin.com/company/mob-avenue/).

  7. Valuations in my opinion are a concern at current levels. IronSrc at 1 Bn revenue is valued at $2 Bn, and they are quiet profitable. Vungle at 300 Mn revenue got bought out by buy out firms at 750 Mn. Mobvista and Taptica are publicly traded and at 2x Revenue multiple. https://www.bloomberg.com/quote/1860:HK
    https://www.sharesmagazine.co.uk/shares/share/TAP
    Some like InMobi, Liftoff, are privately held and at 3-4x Revenue multiples based on last funding rounds. All of the above firms playing in the third party marketplace beyond Google/FB have access to their own direct SDK inventory, bidders/tech stack, Ad Mediation stack, access to exclusive audiences and even O&O properties. Affle at 320 Cr revenue is valued at 4000 Cr.

  8. Positives : The positive in my opinion are their acquisitions of RevX and Vizury, which plays in DSP and re-marketing space. Criteo and other re-marketing players do well and further, advertisers beyond a point will look for re-marketing spends as their user acquisition would have plateaued. But both the businesses are currently small, and further would need significant tech investments to compete against larger players and they need partnerships to gain access to audiences/user profiles which can be exclusively accessed through them.
    Secondly, the market in India/SEA is still at a nascent stage and growing. So they will be able to show growth as long as marketers and firms are hungry for new users, but local competition with the likes of chinese players like Mobvista, YeahMobi, Baidu, Israeli players like Taptica/Appnext, Glispa and Indian unicorns like InMobi and smaller networks like Vertoz, Mobave, SVG, Tyroo, Globale Media etc. will also be there. Players like InMobi/Mobvista are much larger in size as well and more global apart from the tech/market focus differentiation.

  9. Lastly, we are yet to see how Amazon Ads and flipkart Ads are other OTT platform ads can disrupt the scene and share of Ad dollars. The happy part is AdTech and Martech space is bound to grow exponentially and couple of Indian firms in the AdTech/Martech space catering to both India and global markets will do very well and can even be deco-corns.

Not invested and not a buy/sell recommendation. Happy to discuss further on any clarifications on the forum. Hope the data shared is useful.

67 Likes

Quite clear insights on the upcoming quarter results. Given current valuations, any disappointment on the earnings growth will send this stock into a down spiral.

2 Likes

Just a few words of caution. To start with I do not understand very well what Affle does and what does having such large number of user profile or data points mean. But there are a few very standard questions that I try to seek an answer too:

a) Most of the tech companies can use a ton of jargon, the more the jargon used the more stuff a company has to obfuscate.
b) What is being expensed out and what is being capitalized in terms of IP (fancy name for code). A lot of product companies are under the hood services companies. The right way to look at what can be capitalized is to see what has terminal value (very similar to mfg companies). Most of the times the terminal value of almost all IP with a s/w company is virtually zero.
c) Patents: This is another red area. We do see a lot of patent suits and some very quality patents which deliver royalty, creates exclusivity and give a huge advantage to the patent holder. But from my experience over 95% of patents are worthless and that is being conservative. Other than helping a company market itself most patents have zero value. For that matter I become extremely cautious when I see companies who own patents for the sake of owning them.
d) Understanding the competitive landscape is paramount. It is also equally important is understanding the ecosystem. A good question to ask is: If the opportunity size is large why doesn’t a Google, Facebook or an Amazon do it? Another question to ask is if the underlying platforms are Youtube/Google/Amazon/FB/Android Apps can the functionality be provided in the build in sdk/platform etc.
g) Data privacy laws are getting stringent every passing year. Companies relying on user data will have to take explicit permissions in ways to use that data. Any lawsuit relating to privacy can virtually kill a company which specifically depends on user data. Similar scenarios exist for data security.
h) Acquisitions and goodwill. No different from mfg. companies but unless one can understand how does the acquisition fit and why the same could not have been created inside, all acquisitions are duds. The more strategic an acquisition is the more pointless it generally is.
i) Don’t get carried away by the presence of marquee investors but dont ignore the usual suspects. Specifically incase of Affle it would make sense to understand why Bennett Coleman & Co. is a pre ipo shareholder of the company?
i) Our innate inability to understand the fast changing technology world. Microsoft around 12 years back acquired aQuantive for $ 6.3 bn. This investment was written off completely in 2012. This is what MSFT said at that point of time:

After 5 years:

22 Likes

Concall transcript of Affle uploaded on bseindia.com. also details of proposed meeting and concall with several investors like Soros Capital New Vernon Helios and many others investment firms and brokerages also posted on BSE.

CONCALL TRANSCRIPT NOV 19.pdf (436.3 KB)

https://www.bseindia.com/xml-data/corpfiling/AttachLive/85af2ad1-e2f0-4b85-8570-7c45240fc892.pdfPPT 15 NOV 19.pdf (1.6 MB)

1 Like

Key indicator of Affle’s demand drivers – app downloads and installs

Disc.- Invested

1 Like

Institutional holding of affle increases by 0.38% from Sept Quarter.

It also gained inflows from MSCI

In my previous role at a Fintech company have worked with Affele. My observation

  1. they have no data science, Ml, or deep consumer profiles. They can get u cheap installs/CPC but beyond that its completely shitstorm. The reason is simple that no company passes any data back to them so they never which profiles are working or not (unlike google or FB). With no or limited feedback no ML/AI models work

  2. they were cheaper at installs/ CPC but way too expensive for customer in-app action based performance compared to other ad platforms

However, they have good relations with a lot of large Indian companies …and our integrated into digital plans for these companies. And we know it’s really hard for a big companies marketing manager to suddenly shift out a large player

Concluding remarks : product is not great with very limited targeting capabilities

But, they have deep relations which means revenues will keep coming.

PS- not invested

PPS- in next call if possible ask them to break revenues and costs under ad platforms, programmatic and incent

14 Likes

I love your last bit ask the breakup between ad-platforms, incent and programmatic.

I am coming from digital background, work on customer acquisition side with brands, and feel exactly like what your mentioning.

Affle services are plain vanilla which many many companies in digital industry offer, they just have larger scale. if one starts looking one would find atleast 10 companies doing revenues between Rs20-50 crore annually. Markets seem to not understand their model, and for some reason are positively spooked by their “Digital Model”.

4 Likes

Though y-o-y results look good, q-t-q result is not to much promising. Is the business waning?(Disc.: invested)

9M FY2020 Highlights:
:black_small_square: Revenue of Rs. 253.8 crores, an increase of 34.3% y-o-y
:black_small_square: EBITDA at Rs. 67.7 crores, an increase of 34.8% y-o-y
:black_small_square: EBITDA margin up by 0.1% to reach 26.7% from 26.6% in 9M last year
:black_small_square: PAT at Rs. 50.2 crores, an increase of 46.2% y-o-y
:black_small_square: PAT margin up by 1.4% to reach 19.6% from 18.2% in 9M last year

Q3 FY2020 Highlights:
:black_small_square: Revenue of Rs. 94.5 crores, an increase of 27.4% y-o-y
:black_small_square: EBITDA at Rs. 27.3 crores, an increase of 20.0% y-o-y
:black_small_square: PAT at Rs. 21.4 crores, an increase of 31.3% y-o-y
:black_small_square: PAT margin up by 0.3% to reach 22.3% from 22.0% in Q3 last year

the business is cyclical based on quarter. qoq comparisons not correct

1 Like