-
Yes, comparable firms would help. Mobvista trading in HK is a comparable. Millenial Media was a comparison. In the private side InMobi, Applovin and IronSrc give some comparables. Digital Turbine as well in the OEM monetization. Affle is doing well as a company and growing in CAGR, but valuation at 15-20x revenue is like a SaaS/Product firm to what they are actually doing.
-
TradeDesk to Affle comparison is like comparing a product firm ( Microsoft) with a services firm ( Accenture), both are individually good companies irrespective. I also see parallels in how Cupid/LT Foods end up being looked upon as FMCG firms, and here even more due to scarcity of comparable firms.
-
The good part is other affiliate/digital ad agency firms in India like MobAve, Tyroo, Social Beat etc will end up tapping public markets for the valuation premium .
I have personally been working in the ad tech/martech/Saas industry in India for the past seven years. I know a thing or two about affle. While its true that this business and other types of this businesses will continue to grow on the back of high smartphone penetration (this is the main forte of affle and inmobi). However, the devil lies in the details Affle started out as tech oriented company that would give rich media assets to agencies and their respective brands (good margin business). Overtime they started accelerating in what we call ad network or affiliate business especially post 2014 when lot of VC money allowed app based businesses like flipkart, Ola, and Amazon to grow significantly. One of the domino effects it had was that there was a huge demand created to get more users for their apps and the primary channels to tap into these were google and facebook. However, the demand was so big that these businesses started to look for other channels outside Googe and Facebook. This is what gave tailwinds to businesses like inmobi, affle, Vserv, and Tyroo. Affle also benefited a great deal when SVG media group (Tyroo was part of it but then was hived off) got sold to Dentsu allowing only few players to capture the market and affle was very aggressive in wining big relationships especially Amazon India. Even today, majority of the User acquisition campaigns go through affle for Amazon india. This tail wind really propelled affleâs profitability and since then there has been no looking back. They had launched products in mobile attribution, fraud detection, DSP and many others but were mostly cloned from already some of the more established businesses in the ecosystem. With the help of the IPO capital raised, they have strategically started acquiring other tech companies i.e appnext is the most recent one. My only issue here is (speaking strictly from what I have observed with the inside the industry perspective) that their main cash cow still remains the ad network business but now they have amalgamated so many other tech businesses that its hard to separate wheat from the chaff. The market is giving it a premium valuation because of all the tech but I know the core of it still remains purely ad network business and that too because one special relationship it shares with Amazon India. Overall, I am very skeptical of their offerings its true that Inmobi is more of a tech player in this field but even Inmobi has a significant chunk of their revenues coming from ad network business. These business are fine in and itself but in my opinion have too much premium attached to them just because of digital, AI, and other fancy tech words attached to it.
An interview where Soham is responding on dependence on Amazon india
Above was in Feb that top 10 customer contributes less than 50%,
Add to it inorganic growth and global revenue side story, concentration risk would have come down further by now.
In many ways itâs a scarcity premium attached to such companies in listed universe , add to it capital allocation is relevant ( performance to be seen), like the timing and aggression on acquisition as long as it is in same space.
Soham is the key jockey here IMO, guy seem to know what he is doing, tech stack by itself is commodity, how one leverage it to deliver value as a platform/solution at scale is key. One can listen to his listing ceremony speech, outside regular media interactions.
Amazon being a loyal customer is a testimonial to their performance, mobile app eco system and economy has only one way to goâŚ
Happy to ride it along as long as growth runway, capital allocation, trust in jockey and reported numbers are inlineâŚ
Invested!!
On this forum and some other folks I talk to have echoed similar views and I get that it is not a special company as such. What they deliver is still basic and could face serious competition in the future. Then I look at NRN of Infy who created massive wealth starting with body shopping which was considered absolute kachra biz model at one time waiting to be over after Y2K. I recall similar disdain for low commodity grade work but it was the industry growth which saved them and they moved up the value chain along the way. The promoter is smart here and the industry itself is growing very fast so there are many optionalities which canât be estimated today. It is better for investors to keep watching the developments like a hawk.
- Fair comparison, but no disdain for low commodity work. It is around the valuation premium. Here the private company valuations backed by VCs are lower than public market valautions. Infosys value at 2001 got breached only 10 years later, though the company did well
- Amazon UA budgets mainly move to FB , Google, InMobi and then Affle. Plus there is IPG group. mFilterate is the anti fraud partner with Branch as attribution partner and their market inputs would help. Affle works with Amazon to re-distribute the campaigns to other players like Mobvista, China and Israel based ad networks down the line. Flipkart works with these partners directly while Amazon works with 2-3 partners directly and rest through IPG. Would need to be watchful on Amazon budgets move.
- Ad Network as a business is large client dependency. Small players would not bid on digital/mobile oer third party networks, will do self serve FB/Google Ads. Hence that is not a major concern. Even IT Services is large enterprise deals primarily.
- The promoters are smart and a good team, and relatively did better in valuation compared to SVG Group. SVG with similar revenues got sold for < 1000 Cr.
- Will be good to see how they move up the value chain using the cash flows from Ad Network business. IronSrc did a similar move 5 years into gaming ads and now studio business. Their revenues are upwards of $ 1 bn now.
- Tech disruption plays are very difficult to predict, smaller players can move faster to capture client budgets with off the shelf products and as Sumit said, need to keep a close watch.
The advertisers use a unique device ID number called the IDFA to better target ads and estimate their effectiveness. In IOS14, each app that wants to use these identifiers will ask users to opt in to tracking when the app is first launched.
This will impact on the ad inventory and more so because of the acquisition of the app-Next.
I have seen many people using apple devices because of the privacy and security it provides and I feel that this trend is catching up. People are becoming conscious about their personal information being shared by the likes of google FB and other apps. Thus one can not completely ignore the possibility of google enforcing such policies in their coming updates.
Moreover, applying these kind of policies give google an uppeehand considering their very less dependency on other apps for ad inventory. Applying such policies will provide them a monopoly position in the ad inventory market. Same goes with the FB, with people knowingly sharing their personal details on FB itself, facebook is also not dependant on other apps for their inventory.
Disc: Was holding the Affle but sold couple of days back because I donât feel that the valuation is justified.
-
This is a good thread on the topic
https://twitter.com/madnavin/status/1298620343192391684
How a single update might change the face of in-app advertising forever? -
This is how Ad Networks work ,
"Imagine trying to run your ad on 100 different news apps and talking to each entity separately. Itâs inefficient, time-consuming and more importantly, an exercise in futility. Because hereâs the thing â you could get a broker to do it for you.
Enter Ad Networks. They connect you to the news apps or any other app for that matter and help you run your campaign in more place than one. Hereâs how it works.
Think Finshots. If we wanted to monetize our content by running ads, weâd go to an Ad Network â like Facebook and see how we could facilitate this process. Weâd be willing to run ads so long as we can make some money off of it. And Facebook in all likelihood will tell us we would need to **slip in a few lines of extra code in our app that will help them get the ads up and running."
-
Here in Affle case, they have a SaaS platform from Affise/CAKE and acquied one from RevX ( brand campaigns ) to run the campaigns form advertisers on inventory of apps like Finshots through AdX ( Google), FAN ( FB), MoPub ( part of Twitter), InMobi exchange , Mintegral which have their SDK ( lines of code) in all these apps. Affle doesnt have its own piece of code. Further, the user data collected and targeting is also limited as the campaign targeting is done by the intermediaries and not Affle.
-
In FMCG terms, there is a brand like HUL selling products and a distributor uses the HUL system to record data of sales by outlet in a particular area. If HUL does well, Distributor does well. But Distributor does not own the data. He knows which outlets are there in a area, but not the DB of sales by SKU. Here Affle is like a Distributor, and digital Ads growth should aid their growth, but the data control and ownership is not like Amazon or DMart which work directly with the brand and get customers.
-
Will a distributor be valued like a FMCG Brand ? Thats for Mr. /Mrs. Market to tell. Affle can also use the cash it generates and better valuations to raise more money and get deeper in the ecosystem. Owning an app or building its own systems for better targeting etc. This is tough game, lot of execution risks ( consumer apps survival ratio is low) and needs lot of capital burn.
-
I guess this is what we are discussing, macros are right and digital marketing is growing, but the stage at which the company is, the tech capabilities and execution risks should the company be a 7500 Cr company at 350-400 Cr revenue.
Posting my brief analysis.
Disclosure: invested.
But also it should be noted that majority ad spends are through IOS users & apple putting a privacy issue might affect company materially as value wise spends are highest.
It wont affect Affle much unless Google also wants to take a share of Ads money through playstore/android.
Again, I feel there are so many nuances to these technologies that one must take care.
Affle is an ad network but there is another definition of an adnetwork that we use in the industry. They are known as Blind networks, the reason for that is they donât always have control over where the advertisements of these advertisers run and they cannot guarantee any fixed placements. They are known as cost effective and scalable platforms but the quality is generally perceived to be sub- par. So, the question is why do such digital advertisers use them?
Well, every advertiser has limited budget and teams have certain benchmarks, KPI (Key performance indicators ) to meet i.e they need many installs, registrations, leads, etc at a cost effective basis.
If you run your campaigns on Facebook and Google you will generally have to spend more because the inventory they offer is expensive but genuine and you can get to know where they ads are running with all the insights. In the case of ad networks there is no such transparency the reason why adnetworks find space in the media plans of advertisers is because they can offer very optimal pricing and give them way more leads, installs, registration than what facebook and google but the quality is not guaranteed.
The advertiser is okay to make that trade off because the marketing teams need to meet their targets and there fore its essential . This has how the industry have been run for the past 5 years but it is increasingly becoming wary of the quality of these adnetworks and therefore there is a genuine concern on the future outlook of such businesses.
Secondly, the SAAS platform that affle uses affise (Russian based) and Cake (American based) are affiliated based software tool that enables affle to procure traffic /inventory from various long tail publishers to run these campaigns. The other troubling aspect of this area is that majority of the inventory on the affiliate channels come from China. This has been a secret that not many have known but until very recently majority of the digital inventory available indian outside of Facebook, Google, and some native publishers were Chinese and many adnetworks and yes including affle were procuring cheap traffic from them and making good arbitrage when selling to the advertisers.
Again, I have personally been in this business and think there is nothing inherently wrong in the business model but if your story is that of a tech player then its bit troubling to know that majority of your cash flow is still been generated from this rudimentary ad network business.
Secondly, the RevX acquistion was made to improve their remarketing and retargeting product offering. This platform that affle acquired was a niche player in selling their remarketing tool where through cookies based targeting they would re-target users for advertisers that had not bough their product earlier or for some reason fell from the conversion funnel as we call it in the industry.
Now, with google and apple disallowing cookies based targeting advertisers are also increasingly getting skeptical of allowing any third party software i.e. Criteo and RevX to drop their cookies on their website and undertake retargeting campaigns.
So there is a huge question on the future outlook on this business as well.
Now, Let me reiterate I am in no way saying that the management and the company is doing wrong. I think they are capable management and they run a good ad network business that continue to grow in a gradual way.
However, the story that has been told in the market of this heavy tech player is only partially true because majority of it is happening on the back of acquistions in the market which may or may not work out.
Therefore, I still feel that the valuations is too steep and no real predictable way to know its future growth outlook.
Good to see them hiring in current scenario across functions.
Acquisitions - which they have been selectively aggressive about - one lever to increase margins would be to get tech jobs delivered out of india.
Good post.
In AdNetworks, there are those which have SDK inventory (own code in publisher apps like Cricinfo) and others which are blind inventory with Chinese browser/key board apps and no control on creative or placement. Blind Networks are more commoditized and all that is needed is a SaaS platform like Affise , few advertiser and supply connections to start one. Affle is in this space and hence as you said the valuations on par with TradeDesk or Google is a stretch. Also one of the reasons Amazon works with Affiliate/ad Networks in India is the local is averse to spend on Google/FB as that would involve sharing of data. They do re-targeting with Google/FB and user acquisition with InMobi, Affle and other networks. Flipkart on the other hand has majority spends skewed to Google/FB.
One other thing I realised while speaking to a seasoned investor is, companies with low float ( high Promoter + FII+ DII holding), with the support of few funds can spin a âunique in the marketâ story where there are few listed comparables to compare with and the Mr. Market takes the price to astronomical levels.
Hence need to be careful and as someone else said be watchful and clearly ascertain if the tech based play is working out.
Why didnât they opt for a VC , and why go for IPO ?
Lets say you own Affle. You want to raise money. VC or PE offers you Rs100 for the stake you are selling. You realize you can raise Rs110, 120, 130 or 140 in an IPO. Would you raise money at the highest valuation or sell to VC who understands technology better?
Co has Computers of only 50 lakh as on March 2019 (IPO Date) (30 Lakhs Net Block). Are the balance leased. ?
Company gross block of computers is Rs158.6 lakhs end FY20 and Rs90.6 FY19 which will be the cost of buying the computers. The net block post depreciation (If I am not wrong they are depreciated in 3 years) value is Rs75.2 lakhs end FY20 and 53.4 lakhs end FY19.
Total number of employees end FY20 is 295. Rs158.6 lakh/ 295 is Rs53,762 per computer which is a very decent price per computer.
Please note the cost of computer software (amortised over 5 years) is seperate at Rs251.2 lakhs end of FY20 in intangible assets. Also note there is software application devlopment in Rs2525.6 lakhs which 10x of computer software.
Bulk of Cost is Inventory/Data costs - Normally in IT cos, employee cost is 50%+
Employee cost in FY20 is Rs2729.3 lakhs. Number of employees 295. Per employee cost is Rs9.25 lakhs. Dont see anything fishy in that.
Affle is not an IT company. It is involved in mobile adverstising. You need ad inventory to do business at it is the single biggest cost.
If I get answers or insights on some of your other questions I will get back with more.
Disclosure: Invested, but always ready to hear positives or negatives to retest investment thesis.
-
If IPO is better pricing than VC , you will surely go for IPO. But why is the case that neither Tyroo/SVG or Affle or other Ad Networks like MobAve etc. not attracting any VC/PE interest. A case of âWhy the dog didnt barkâ ?
-
âAffle is not an IT company. It is involved in mobile adverstising. You need ad inventory to do business at it is the single biggest cost.â - Right, so why compare with TradeDesk ? Should the actual revenue be counted post inventory costs ( Net Revenue = Gross Revenue-Inventory costs as it is a pass through ?
-
Avg. employee cost at 9.25 L, engineering cost is higher. Fresher salary at Redbus or Cisco is 9L minimum
-
Computer cost, not sure what is the number post depreciation but developer laptops ( 16 GB RAM) cost more. iPad Pro 13 inch is 1.2L
-
They buy third party software ( Affise/CAKE) and would have cloud infra costs as well for RevX. Need to check the same.
âAffle is not an IT company. It is involved in mobile adverstising. You need ad inventory to do business at it is the single biggest cost.â - Right, so why compare with TradeDesk ?
I did not compare it with TradeDesk.
Should the actual revenue be counted post inventory costs ( Net Revenue = Gross Revenue-Inventory costs as it is a pass through ?
Compare it the way the you want. Does not change the EBITDA or profits.
But why is the case that neither Tyroo/SVG or Affle or other Ad Networks like MobAve etc. not attracting any VC/PE interest. A case of âWhy the dog didnt barkâ ?
Why are you assuming they did not attract VC/PE interest. Isnât it possible these companies were approached but were not happy with the valuations on offer.
Avg. employee cost at 9.25 L, engineering cost is higher. Fresher salary at Redbus or Cisco is 9L minimum
Are they overpaying or underpaying? If they are underpaying then it is inline with your arguments it is not as high end tech or AI.
This kind of reminds me of a company called Crompton Greaves. Used to make high end transformers and switchgear. Kept focussing on the high tech business and paid less attention to low end business of making fans, lights, appliances etc. Finally the so called high tech business went belly up, but so called low end business fetched handsome valuations when promoter sold to PE.
Computer cost, not sure what is the number post depreciation but developer laptops ( 16 GB RAM) cost more. iPad Pro 13 inch is 1.2L
Which software developer or coder works on iPads? I am assuming you typed that by mistake. No Indian company buys apple products. You can buy extremely powerful laptops at 70-80K. As corporate when you buy in bulk you can get bulk discounts also.
Thanks for bringing these observations to light Ashwini.
.The IPO came in Sep 2019. Till July 2018, it had no subsidiaries etc. Infact they started acquiring a lot of cos just 1-1.5 year before IPO and today those cos contribute 50% by way of turnover and profits.
**This is partially why I was very uncomfortable with their story. I know for a fact that they were purely an ad network until 2017. In fact a little trivia, they had at one point considered going for IPO in USA selling the Indiaâs internet story but then decided to simply do it in India. Its really curious how they went about on an acquisition spree right upto the IPO strengthening their story and continue to amplify their revenues through different businesses as they really didnât have their own IP.
Microsoft and D2C didnt invest in the company for current business
-
- Microsoft purchased shares in 2009 (even before Internet boom began) - for its SMS platform
-
- Similarly Japanese CO D2C Inc
This is also correct, as I said they have gone through many reincarnations and they were earlier simply known as rich media provider company supporting major agencies on creatives for their clients like Samsung.
- Similarly Japanese CO D2C Inc
- I was referring to earlier discussion on comparison with TradeDesk and its P/E as a Ad tech SaaS firm.
- Fair point, though the general practice in agency/media business is net revenue. The profits dont change, only multiples will look different.
- The point I am making is , if it is a outlier tech firm with amazing execution it is highly likely a VC/PE would have provided better valuations and preparation for IPO listing in India/NASDAQ.
Maybe I am mistaken. - The avg. salary is low, but the employee split by engineers and new joinees is not known so cant see this clearly. Hence need not put weightage here. The low end or any business with excellent profitability deserves valuation, and guess we are discussing if the valuation multiple is justified at 100 or lower and the P/E comes from the moat which generally low end businesses do not command. There is significant competition as well, just that they are not listed. Sokrati does close 900 Cr in media spends and good profitability but sold for much lower revenue multiples.
- I meant iMac Pro, which and other 16 GB RAM machines are used by developers and have been sourcing laptops for my own startup and hence aware of the pricing. Any case without knowing the split of infra/cloud costs, SaaS product costs by RevX/CPI business it is difficult to give clarity here and hence again a weak point to discuss. As a side note Indian product companies buy a lot of Apple products especially for developers , except if you are in Microsoft
I think someone from a tech IT team need to jump in to answer these.
Just to put in a perspective, I am working in a US based MNC, they provide Mac Pro ~2L per person, earlier I worked in an Indian IT company, they were aprox 200 sized team, they provided desktop worth 40K but these were the cheapest, rest had laptops worth 60K and even IT companies has inventory of PC incase one goes bump, so I would average 60-70K!
This excludes
Just my experience