Buying Efficiencies : Continued ... CPC (Cost per Click), digital marketing

Topic Continued : Media buying efficiencies : The Great CPC media...

I know this took a lot of time to come thru, but thrust me it will be worth the wait(weight for some).

So here's my take on getting efficiencies into CPC purchase. 

When you buy CPC in the market today, this is what it is sold as. 

1. It is the most effective media purchase.
2. It is like buying user engagement.
3. It is user engagement.
4. The user has better intent.
5. It's like buying on Google content network (Hehe Haha)
6. you just cant go wrong on this media currency purchase

Some of the things said above are to some extent true, but don't follow them blindly. 

The Catch 

1. Most of the media planner buy inventories of publisher reporting ( It is changing but not completely).
2. The Tech involved in tracking clicks on the advertiser side is non existent ( it is changing but the tracking used is crazily ineffective.)[I will explain this later]
3. Analytic are completely missing... post click or even the click itself.
4. There is this absurd rule prevalent in the market that click bought will result into visitors...[ Again I will explain this later]
5. Most of the clicks bought are on blind networks.... interestingly they do some crazy business in India.

So here is the explaining and how you can avoid yourself from falling into one of the sales pitches mentioned above and bring more efficiencies to your media plan. 

CPC is a good inventory to buy for sure, no doubt about it. It is cost effective, reliable and can be scaled up or down based on your media spend requirements and communication objectives. Having said that, please follow the some of the below mentioned rules of engagement while buying CPC inventory. 

1. Know the source of inventory. lots of time the planner does not know where in the world are clicks coming from .
2. Get basic analytic code implemented on the click tag " UTM" parameters, if you are using google analytic for tracking
3. Serve banners where ever possible. Use good efficient technology to serve ads directly into any ad network. It might "add" a few RUPEEEESSS to your budget but trust me it will all be worth it in the end . Technologies like "ZEDO" can come handy, if you don't want to invest in double-click and also get Indian support 24x7 , there are a few more... let me know if you need any help here.
4. Get your ad serving partner to implement click fraud detection to insure best buys and no fraud happening on the campaign.
5. Get your tech partner to do the correct geo targeting and not get clicks from unwanted territories. 
6. Also if you are also using some attribution technology pls drop cookie to bucket users on to the site, enabling you better tracking over a period of time and also measure the quality of the traffic. 

After you have taken corrective step while buying, keep a strict monitoring on the campaign when it is live to ensure that all that was planned is being delivered on the dot. 


Please keep in mind that a few things will go wrong, specially in India on CPC inventories.

1. Amount of clicks bought will not reflect in similar numbers on the analytics. there are a few reasons for this some true some myths... a separate post for that for sure.
2. Geo targeting will be off the mark by 10-30%. i.e. 10-30% of the geo targeted area will not be able to see your ads so less clicks and 10-30% of people out side of your targeted area will see your ads, so wastage.
3. Not All CPC networks / publishers will let you run ad-tags. So buy from the people you trust and not some fly-by-night site operator etc. 

Also use google analytic as one source of information and not the only truth. Use a parallel tech, or simply use server log reports to compare google analytic for a better insight, or use a "paid tool".

Also to note that buying CPC on networks or mainstream publishers is not like buying on google content network. In google, they have a lot of tech going into targeting, fraud detection, measure-ability, real-time bidding etc.

PS. New Kid on the BLOCK : COST PER VISITOR.Who says "Don't buy clicks", buy visitors. (Pls do not confuse this with something called as FPV or Full page views)  more performance to the advertiser... Interesting.. right.. 

Next post : Cost per "ACTION"... that's where all the action is... will be posting an undate on CPM efficiencies also , do read. 

Digital Traffic Attribution & User tracking : A Basic Framework

Schematics of the Attribution technology:  on digital marketing.

The last time I wrote on attribution. Got a few requests on how the attribution technology should work. Thought I should BLOG it down for people to develop own technology, rather than hiring it at exorbitant pricing.

Things required before getting to develop the product.

1. Knowledge on Cookie, super cookie.
1. Cookie length Vs Cookie Validity
2. Referring URL
3. Traffic Quality 
1. Source 1 Vs Source 
2. Paid Vs Free 
4. Business Rules
1. First Referrer vs Last Referrer
2.  Many referrer 
5. Multi Channel Digital Channel 
1. Media 
2. Search 
3. Affiliate 
4. Mailer 
5. Social
6. Direct

Why is attribution so important on the multi channel digital marketing strategy. Some of the reasons for this are as follows.

1. Large media monies are being spent to get customers on the online space for
1. Brand popularity
2. Transaction
3. Reach
4. Frequency
5. Top of the mind recall etc etc
 
2. To provide the logical information to marketer about what channel is working
3. Performance is the key for a digital marketing campaign to be measured.
4. To know what worked best for the brand/ product / service on the digital channel.

One of the most interesting things to know is the fact that most companies that are in the eCommerce space in India don’t have this in place and don’t seem to have any possible solution to develop it.

Issues a marketer faces in attribution if the technology is not good are as follows:-

1. Multiple refers show us as claimants to a particular interaction
2. Allocation become an issue if the traffic provider is a paid one
3. User journey cannot be defined and hence measured at various steps
4. Budgeting for next campaign becomes an issue as you don’t know what worked the best for you.

So here goes my attempt into developing a schematic to a digital marketing attribution/ user tracking technology.

The landscape 

The advertisers landscape for attribution

The user Journey


Stage 1 : Research. 

The main thing to notice over here is the fact that the user might choose to engage with the brand using many channels or only one, but for a fact he will use two or three channels during his research work before buying a product . So it is important to allocate this user initially to the source of his first visit. Then with the same cookie id keep on updating his profile, add various data points like IP address(this way you can build the users profile over a period of time, city, service provider etc etc.)  


Now once the users moves from this bucket into the consideration set, you can add the  following data to the same cookie id. 


Stage 2/3 : Consideration / re-evaluation 

Product being the main, keep on adding the products that he browses thru, along with date &  time of the visit (again adding more to his user profile). Within this set of data, I am sure you would also capture the source of the user, the referrer, the publisher etc.
The source data / referrer data will be helpful later when we will look at the sales attribution. (Which the topic is all about). 

Stage 4 : Purchase  

In the sales stage of the user journey, just add the product purchased data to his profile in the same cookie id.


This is important as this is also one of big mistake large eCom companies do when they are doing cookie tracking on repeat visitor, they continue to show even the purchased product in banners when the users is being re-targeted on re-targeting networks out side the site or they even keep on throwing up the same product in the consideration set when the user is on the site again, thus adding nuisance value to the user's website journey.


Store the information both ways, If the user has purchased the product then store all the cookie id information into his login id., for future reference and his loginid details into the cookie id, enabling you to target or re-target him better. Next time the user pops onto your site or on your re-targeting network.


Important thing to note over here is that users who either delete cookies or come back after the expiration of the cookie validity period can be found if they again log into the system, enabling advertisers to offer a more customized product offering to them. 


Coming back to attribution of sales to the media channel.


Now that we have the details no the user and the road he used to travel to the site.

Now based on the business logic. first referrer wins or last referrer wins scenario you can attribute sales. 

Also if you organization believes in fairly compensating all  the parties in the user journey then also this data set would help in doing so.

This technology if developed can enable the advertiser to also do the following.

1. Fire tracking pixels of media channels based on business logic.
2. De-dupe sales data across media channels.
3. Find effective data for more efficient buying on the digital media.
4. provide some piece of mind to its eCom managers, doing Multi channel.

Will be happy to receive your feedback on the same, or if you feel I have missed out on some thing(s) please send me a mail or comment, will try to help and resolve your attribution issue.