Archives for posts with tag: conversion attribution

Automated bid management software is becoming more and more popular and the technology and algorithms powering it are becoming more advanced.  There are a number of providers who have built extremely powerful software solutions for automating PPC bid management and the functionality they offer is immense.  But I really do not believe they can ever fully replace human management of pay per click campaigns, here are few reasons why.

Conversion Attribution – Beyond CPA

Conversion attribution is something that I believe is at the heart of a truly integrated digital marketing programme.  The ability to allocate conversions and contribution to conversion across multiple digital channels will revolutionise the allocation of digital budgets once a truly accurate method of measurement is found.  It has been the principle of multi-channel digital campaigns for a long time, build brand awareness through display which will lead people to the search engines and conversion.  But very few, if any, tools exist which can effectively prove, more importantly, accurately report, the affect each channel has had on an individual conversion.

Automated bid management software analyses individual keywords in isolation and doesnt account for the impact a click on a non brand keyword might have on a brand keyword for example.  By managing each individual keyword at a set CPA you arent giving them the credit they may (or may not) deserve and you could be seriously limiting the volume of leads available through PPC.

Critical Mass and Sales Volume

Profitability and return on investment is great, and it is the main advantage digital channels and PPC have over our offline counterparts, but it is not everything!  Many businesses have a critical mass of leads or sales they need to keep themselves afloat, I deal with them all the time.  And whilst they buy in to the accountability and measurability of PPC if the phone stops ringing, or the leads stop coming in, this all goes out of the window as their business depends on a certain level of sales each day.  Rule based bidding is useless in this situation and you have to go after the volume.  This is especially prevalent in finite markets where a large research process isn’t undertaken.

Variable Conversion Value

Each conversion is not necessarily worth the same as another, and it is difficult to accurately provide an individual value for each PPC conversion.  In this way it would take a lot of human analysis and intervention outside of the software to manage and maintain profitability.  This often occurs when the online action is an enquiry which results in an offline sale.  One online enquiry might have a resulting sales value of £20 and another might be £2000, in this situation, how can you effectively provide a profitable CPA for these enquiries?

Offline Conversions

Offline conversions happen in every market, telephone numbers on site, visitors to a store, customer service question which result in sale.  Automated bid mangement software can never effectively account for these sales and so will managing the campaign with only half the information.

So what is the solution?

Firstly, I believe automated bid management software has a part to play in managing large and complex PPC and Adwords campaigns.  But there needs to be a large amount of human intervention and management which leads to the statement made in the title, human management can never be fully replaced by bid management software, so how is this best managed?

One option is to employ the 80/20 rule of bid management.  By allowing a software package to manage the 80% of keywords (the long tail) which generate 20% of sales you cand remove a large burden from campaigns with thousands of keywords and focus you attention on the important 20% of keywords.

Another is to use the human element to analyse the offline and supplemntary data which cannot be interpreted by the software and build complex and evolving rules for the software to follow.  I know some of the software providers are frustrated that many of their customers only us a small percentage of their tools functions and no-one really pushes the boundaries.  By using the human element to analyse the reams of offline data, plug it into the tool, or learn the most effective ways of managing things in an automated manor you can ensure that you:

1. get the most the software has to offer
2. fill in the gaps and analayse the data and factors the software is not aware of

In truth, I dont know the most effective solution.  I am sure software is going to play an increasingly important role in PPC and digital marketing but it can never replace the human touch.  So if you are planning to use one of the available tools, or your agency are going to do so, make sure you dont rely solely on technology, it will never know the full story.

I have been reading up recently on tracking solutions I came across the argument between first click and last click tracking. More specifically which one gives the most accurate measure of performance?

First click tracking is the term given to solutions which cookie the user on their first visit to a website and record the source of their visit.  On each subsequent visit they are recognised as the same visitor and allocated back to the original source and so when they convert through to sale/application/sign up the software records them this as the origin of the conversion.

Last click tracking relies on session cookies and so no matter how many times a user visits a site, the source of their conversion will be recorded as the final means by which they reached the site.

So which is a more accurate measure of advertising performance? Well its difficult to say really.  Display advertisers will normally put the argument across for first click tracking as they play the brand card a lot more and place their strength in building awareness of offers and products.  Some of them even cookie a user on ad impression so would even more stringently argue this case.

Search advertisers on the other hand are more likely to lean towards the argument for last click tracking as users will often use search results to find brand and products they already know they want to buy.  Regardless of where they found out about them a user is more likely to visit Google and search on a brand or product name to complete their purchase.

I’m more likely to not worry about it and assess online as a whole rather than worrying about each channel in isolation.  More and more recently I have noticed the effect the channels have on each another and I am now suggesting the holistic approach to online is the best way to go.  Conversion attribution is a difficult skill and requires a highly sophisticated piece of tracking software.  By pulling one channel you could quite easily find that although when analysed in isolation it didn’t appear to work, it was having a positive affect on another channel and you see a drop off across the board by its removal.

My advice? Although all channels need to be measured, pay careful consideration to the impact they may be having on one and other.  Look at online as a whole to measure performance before making any decisions as the halo effect exists and if you pull one brick from the tower, it could all come tumbling down!