So what in the world happened?…315!

For those who don’t know what 315 represents, it stands for March 15, which is “International Consumer Rights Protection Day.” Now every year on March 15, CCTV airs a two-hour long special, which exposes companies that violate consumer rights. The program covers various issues across multiple industries, such as McDonald’s food safety in last year’s program. But this year, the knife turned to the digital marketing industry.

It all began when the handsome CCTV host said, “There’s this group of people that you do not know, but they know everything about you including: where you went, what you bought, how many people in your family, your income, what you visit when you go online, what you searched for, etc. ”

For a second, I thought CCTV somehow got hold of my pitch on audience buy  But my heart sank as I realized the gravity of the situation. CCTV, the biggest government-backed media is addressing the issue of Internet consumer privacy. Is the government going to further regulate this industry like I had predicted in my 2013 media buying trends column? The program goes on to talk about the business model of prominent audience buy companies in China, i.e., Ad China, Yoyi, iPinyou, many of which were covered in my previous article. The biggest blow to audience buy was when CCTV stressed these companies made no attempt to let consumers know that their browsing behavior was being tracked. And if the consumer wants to opt out, then the only thing that they can do at the moment is to delete or not accept any cookies in their browser. The entire video can be seen below: (http://v.pps.tv/play_35D6Y1.html)

Immediate ripples throughout the digital media industry

Right after the “expose,” many audience buy companies such as AdChina posted official responses on their company weibo addressing how they deal with consumer privacy issues. Even major online portal companies such as 163.com released statements on how they gather and use consumer information. On social media, various tech KOLs (key online leader) such as Lee Kaifu (former head of Google China) have also addressed the need for an opt-in by consumers for tracking. From a consumer perspective, I pulled together some data on Baidu Index:

If you look at the data, the search terms used were “cookies” and “can cookies be deleted.” Now you can definitely see the spike in search volumes right after March 15, jumping to over 7,000 on that day. This certainly proves that consumers do see this as an issue of concern and they are taking actions themselves to prevent tracking.

Some potential implications

So what are the post-exposure implications to us China marketers? I see that this can go one of two ways:

1. The positive outlook, in addressing this issue will let more people know about audience buy and cookie-based targeting, thus understanding its value to both consumers (personalized content) and advertisers (accurate targeting). In essence, CCTV gave us primetime ads for free!
2. The whole thing goes down south and the government imposes new regulatory policies to limit third-party tracking as well as cookie collection.

Now if it does go down south, there will be some major implications to the China digital media industry and audience buy companies. Specifically if the policy limits cookie collections or requires consumer opt-in, then say “farewell” to the following:

Bye bye!

1. Web analytics
2. Ad frequency capping
3. Re-targeting
4. Media performance tracking (impressions, clicks)
5. DSP, SSP, DMP
6. Taobao/Baidu Union
7. AdChina, iPinyou, Yoyi, etc.

The death of audience buy?

In the end, back to the issue this column is set out to address. Is 315 the death of audience buy? Well, there hasn’t been any immediate reaction by the government yet. And seeing how McDonald’s is still happily in business after last year’s 315, I’d say we’re safe. But since my crystal ball has decided to malfunction on this question. I do not know for sure what’s going to happen in the months to come. But whatever it is, it’ll certainly be a major boost to the China digital marketing industry…or a day that an entire industry comes crashing down.

This article was written for ClickZ

The old way focused on reaching the target audience thru publishers, hence the term media buying. But the digital eco-system is enormous and ever-growing, and consumers are continually seeking content across multiple publishers. For cost reasons, we obviously can’t buy all those publishers. So here’s where audience buy comes in, because what brands care most about are the digital consumer behaviors across different publishers. Hence our focus in audience buy is placed entirely on buying the “right” digital behavior based on various targeting technologies and cross publisher ad serving.

Now all this sounds great conceptually, but is audience buy really a one size fits all solution for brands? Is programmatic audience buying really going to kill off direct media buy? Well, let’s take a closer look at the audience buy model and evaluate.

Media Objective Evaluation

First let’s evaluate if audience buy is really the right buying model for your brand. We can determine that by looking at the primary digital media objectives of your brand. From an industry perspective, FMCG brands are primarily focused on branding across mainstream digital channels; Finance, Insurance, and many B2B brands are focused on lead generation thru some sort of a registration process.

With branding as a primary objective, FMCG brands are the least suitable for the audience buy model, because most of these brands have traditionally always worked with large mainstream digital channel such as portals and online video sites. Now these brands will still want to maintain a direct relationship with publishers for PR bargaining power, this is especially true in China. The other reason being that lots of publishers in China are realizing the growing popularity of audience buy, hence they’ve invested a lot in various content cooperation efforts in recent years. Content cooperation can range from making a micro-film to creating a brand sponsored column or BBS activity. Publishers will often bundle in display ads with content cooperation, so brands still see value in direct media buy. But that doesn’t mean that FMCG brands can ignore audience buy completely, it can still serve a purpose in campaigns where budget is limited and the brand wants to support the main communication channel with a mixture of different targeted publishers.

However, lead generation brands are very suitable for the audience buy model. The primary reasons are these brands’ focus on accurate targeting and action oriented result. Many B2B brands have a very niche target, for example construction, occupational safety, or IT hardware. While many of them do conduct direct buying with category verticals, the more effective lead generation channels have always been performance based such as SEM. Hence, the benefits of using audience buy for B2B brands are two-fold: wide targeting options, and cost efficiency. Traditionally B2B targeted their consumers online thru buying only category verticals instead of large mainstream portals. But that severely limits ad serving impressions because it only catches consumers when they are browsing the vertical site itself. With audience buy, we can target these niche consumers even when they are browsing portals and other non-related websites based on their historical browsing behavior. From a cost perspective, most publishers charge on a CPM basis, but what these B2B brands care about is not impressions but leads. Hence, audience buy can utilize CPC or even CPA method for the long term. This places the focus entirely on the action and sales lead instead of paying for displaying the ad.

Audience Buy Options in China

Now that we evaluated if audience buy is right for your brand, let’s look at some different buying methods in China for you to consider. I’d like to categorize these buying methods primarily by targeting methodology:

1. Media List – this is one of the simplest ways for targeting, which is thru a list of publishers. This method is usually used to target multiple relevant mainstream category verticals covering fashion, food, automotive, and IT industries. However, if your target audience is very niche and won’t be found on mainstream verticals, then consider some of the below methods.
2. Contextual – this targeting mechanism is based on setting a combination of keywords that resembles the brand’s target audience, and then serve the ads if the keyword appears on the website content. This method is usually used to broaden the brand awareness by extending the reach to consumers that are actively browsing relevant content.
3. Search History – this targeting mechanism works by tracking a user’s browsing history in the search engine based on cookies then serving the relevant ads that matches user search intent. Both Baidu Web Union and Google Display Network utilize this method

4. Transaction History – this targeting method rides on the booming growth of eCommerce in China for the past couple of years. Since Taobao is one of the dominant B2C eCommerce channels in China, many Taobao and Tmall brands utilize this type of ads to boost their eCommerce sales. Because if the consumer have bought similar products before, then ad is very relevant.
5. Cross Media Browsing History – this form of targeting is the most sophisticated using a combination of a DSP, SSP, and DMP platform. All the targeting data are gathered from cookies across ads served on multiple publishers. These cookies are then analyzed and each individual consumer cookie gets numerous interest and industry tags. Hence ads are served when the cookie browsing history matches what the brand desires. Now in western markets, the DMP platforms that stores cookies are relatively mature. But in China, it’s still in its infancy and most of the DMP platforms have evolved out of ad network vendors and other ad serving vendors.
6. ISP Browsing History – this targeting method is mostly likely unique to China as in western market it would be considered an infringement of privacy. The mechanism works by cooperating with ISP on entire user browsing history, and then the ISP’s serve the ads thru a pop-up that non-native to the publisher. The ads will load even before the website, because the ISP intercepts the request and serves the ad first.

So as you can see from the numerous interesting targeting methods and companies, audience buy is certainly a growing trend in China. But before jumping on the bandwagon, we must closely examine our own digital media objectives. Then make an informed decision based on what’s really important to our brand.

This article was written for ClickZ

1. Slow adoption of DSP/Ad Exchange inventory buying

DSP (Demand side platform)/RTB (Real time bidding) is a fast growing buying method in Western markets during 2012, and China has also followed up on the trend with both foreign (DoubleClick AdX) and local players (Taobao AdX, Tencent AdX). However, this form of media buying has NOT gained as much adoption as the West. The main barriers to adoption are primarily cultural, as most advertisers will still want to deal directly with influential portals and verticals for relationship and PR bargaining power. For more analysis on DSP hurdles in China, checkout MEC’s head of Interaction Karen Ho’s great analysis here.

 2. Uncertainty for cookie based targeting as the Do-Not-Track (DNT) initiative emerges

Do-Not-Track is a HTTP header field sent by the user’s browser to request certain web sites to disable ANY tracking for that user. Now IE10 and several other browsers (Firefox, 360) have already followed this initiative by setting this option as default for users. However, this initiative is not very well understood in China and media industry bodies have not implemented ANY sort of standards or regulations yet. So right now, media can still ignore the DNT signal as they please. But IF any government regulation is to come through supporting do-not-track, then it will be an absolute killer to any form of cookie based targeting such as: ad serving based on user browsing history, ad frequency capping, as well as re-targeting.

3. Online TV channels moving beyond pre-roll and gets into content production

2012 was a phenomenal year for online TV growth. According to iResearch, online TV industry in China as a whole has nearly doubled their revenue and is projected to reach 17.9 billion RMB in 2013. Pre-rolls will still be the most popular form of media buy on these sites. But online TV channels are moving more and more into becoming original content creators as opposed to purely relying on user-generated content. Starting as early as 2011, Youku have made numerous popular original mini-movies and series from their own production teams, namely “Old Boy”, “Father”, “Miss Puff”. And brands such as Yiliand Nestle have also co-operated with Youku to create their own brand sponsored original mini-movies. As we move in 2013, overall online TV consumer consumption is growing even faster. So it will certainly be another strong year for the online TV industry, and advertisers should look for more innovations and content collaboration/co-creation opportunities with them.

4. The “search war” begins as 360 goes head to head with Baidu

Let the search war begin! 360 launched so.360.cn late 2012, and from the latest Experian Hitwise data, it’s reported to have as much as 10 percent search market share. Now Baidu still dominates at a whopping 60 percent market share, followed by 360 and other stagnant local players like Sohu, Soguo, and Tencent Soso. 360 has gained stable market share in an astounding fast pace due to its popular anti-virus software and browser. And it’s moving fast to commercialize their SEM offerings with a Google partnership as well as copying Baidu branding SEM products like the Brandzone. (Portrayed below) So the search industry in 2013 will certainly be a battlefield with Baidu and 360 duking it out through various means.

5. Social ad effectiveness in question causing a move from display to KOL (Key opinion leaders) based media buying on SNS (Social networking site) channels

This shouldn’t come as a surprise to anyone as Facebook in the West also came under scrutiny for the lacking performance of its social ads. Sina Weibo have been under tremendous pressure to monetize the platform since 2011, and it became one of the top channels for advertisers in 2012 with display ad banners having to book 2 months in advance due to brand competitiveness. But numerous campaign benchmarks have shown that Sina Weibo display ads have absolute abysmal CTR rates, averaging less than 0.5 percent. And the buying methods are NOT based on CPM or CPC, but bulk by month. Hence advertisers are quickly realizing a more cost effective way for media buy on the social channel, which is through KOLs. Popular KOL based media buy channels like weiboyi.com are entirely performance based and can fulfill CPC or even CPT type of buying that eCommerce channels often use.

 6. Mobile media buying on the rise, but would need to integrate with other digital channels

No one can deny the prominence of the mobile channel in 2013, but from a media buying perspective this channel has always been a loner within the media plan. This is largely due to the limited media buying methods on the mobile channel, as display banners and app content placement are the only choices. Often, the mobile campaign consists of primarily a HTML5 page with in-app banner ads to drive traffic. The problem with this format is that it’s totally disjointed with the other online channel’s engagement activities and brands are far from ready to make mobile as the center of the campaign. We have seen some great usage of mobile marketing in the app collaboration space during 2012, especially with some Wechat based campaigns like Nike. As a whole, the mobile ad market is still in its infancy in China, and 2013 would surely see many innovations.

This article was written for ClickZ

1. How is Momo different and unique amongst the other mobile social platforms in China?
2. For me as a regular consumer, how has it using it changed my life?

Campaign Planning

As I ponder upon these great questions of our time, I realized that merely using the app and providing an opinion is NOT enough. Instead I have to truly immerse myself in app thru actual experimentation. With that goal set in mind, I started planning my Momo “campaign”.

Now at the start of every campaign, I’d need to be crystal clear on two things:

1. What’s the overall objective of the campaign?
2. How am I going to evaluate the results?

On determining the objectives, I took the liberty of surveying some of my friends on the “objectives” that they have while using Momo. The insight that came back was that 75% of the respondents say they use the app to “make short-term friends”. (Sample size < 5, male skewed) Being consumer focused, I cannot turn a blind eye to the truth behind their usage. So I decided my personal experiment would also have a makeshift objective of “making short-term friends”. And being a devote performance marketer, I’d measure the performance of the app on reaching this objective.

Setting Measurement KPI’s

But how exactly was I going to evaluate the results? First I would definitely need to benchmark it with a similar app. And second, I’d need to set clear KPI’s on what data needed to be gathered and analyzed. From a benchmark perspective, WeChat is an obvious choice due to relative similar feature sets as well as user base. Now KPI’s on the other hand, would tie closely to how the campaign would be executed. From an execution flow perspective, first messages would be sent to various folks, and the most crucial next step is to get them to actually respond. Then thru my wily charms I have to engage them in friendly conversation, and lastly I’d pop the “invitation”. At the end I’d certainly be looking at metrics from multiple dimensions such as response, engagement, and conversion rate. Let me summarize everything with the below side:

Campaign Execution

With the hard part’s out of the way, let’s get into the fun stuff. For the experiment, I’d send messages to 50 people around me with the same “Creative” and “Copy”. The creative being a nice shot of my abs 5 years ago while in my prime. See below:

Sorry, just kidding… that picture came from the internet. Now for the “Copy”… that’s a hard one as I need to elicit the most responses from strangers. But after searching for “陌陌约炮秘籍” on Baidu, I came across numerous credible sources to guide me on this glorious quest. The conclusion for the best copy was the following: “头像是你么？好像在哪里见过你。” (Is the picture really you? I thought I’ve seen you somewhere.) Here are some of the more hilarious encounters that I had with Momo:

Campaign Results Comparison

Hope you got some good laughs out of those. Now back to the task at hand. After experimenting on Momo, I fired up Wechat and did the exact same thing, with the same number of people contacted as well as the same “Creative” and “Copy”. So let’s look at the results below:

From this analysis we can see that that the overall response rate is slightly lower in Wechat. My initial thought on that is most people use their WeChat ID’s to talk to close friends and colleagues, and some may be reluctant to use it as a tool to make “friends”. The engagement rate however is pretty much the same between the two platforms; I was actually quite surprised by the high engagement rate of these total strangers. But that’s probably because I’m already outside of their primary target audience demographics. Lastly on conversion, both platforms scored poorly with only one acceptance in Momo of my “invitation”.

Post-Campaign Analysis

So to answer what I’ve been wondering from the start. How is Momo different than other mobile social platforms? Well, from a performance perspective, I’d say that Momo is slightly better than WeChat in “making short-term friends”. BUT of course to produce a more convincing and conclusive answer, we’d have to increase the sample size as well as conduct some A/B test on the “Creative” and “Copy”. After all, the platforms are quiet similar in feature set as well as target audience. But how people use it as well as what they use it for will greatly determine their experiences on the platform.

On to answering the second question, how has Momo changed my life? Well, on the short-term, that’s a 4 hours down the drain today talking to absolute strangers. And more it importantly it scored me a night sleeping on the sofa largely due to the below analysis:

Conclusion

On the more serious side, I think that social LBS platforms like Momo are great in breaking down the barrier of meeting new folks around you. This is especially useful in 1^st tier cities like BJ/SH/GZ, where there are lots of out of towners and lonely people looking for new friends and company. From a cultural perspective, my generation of the post 80’s are still quiet eerie about meeting people from digital channels. BUT the posts 90’s are an entirely new breed, for them the line between real-life and digital seems to blur greatly. So as we move into a more digital and mobile focused era, I’m sure many new tools like Momo will surface out on the market. It is important for us marketers to understand the cultural and human aspect of their success, and utilize those insights to better shape our own campaigns.

Now web analytics is nothing new to digital marketers. In fact, most brands and agencies have been doing web analytics for years. If you’re one of those marketers, then ask yourself honestly the following questions:

1. Do you know the purpose of looking at certain metrics?
2. Do any of your web analytics reports tie to your business objectives?
3. Do you gain any insights after reading your GA or Omniture report?
4. Do you know what actions to take after you read your reports?

My guess is you answered no to more than one of these questions. That’s because most companies treat web analytics as a reporting tool, while it’s really an approach to accountable and data-driven digital marketing. A complete web analytics implementation should fundamentally change the way digital marketing is done in your organization.

According to the Web Analytics Maturity Model below by Stephane Hamel (http://blog.immeria.net/2009/08/overview-of-web-analytics-maturity.html), a clear direction is needed in all the following areas for web analytics to be successful: Management/Governance, Objectives, Scope, Team & Expertise, Process & Methodology, and lastly Tools & Technology.

Before I talk about each area of the framework, let me first show you an example how most companies approach web analytics. Consider the evaluation below:

 Now what happened here? The marketing director probably went to some digital marketing conference and saw the wonders of Omniture/Webtrends. As he gets back to the company, he mandates the digital team should immediately evaluate and purchase a web analytics tool. Hence the focus is all about how to choose the BEST tool, and how to magnify scope by tagging anything and everything under the sun. But as this company rolls out the shiny new tool, no one is clear on the business objectives of what and why their tracking. Since it’s a technical tool, the IT team gets slapped with weekly reporting duties. But the IT team has no clear methodology in reporting the right metrics for the marketing team. And most importantly, no one is clear on a governance model on how insights and recommendations will be executed. As a result, this company’s web analytics effort ends down the road of data puke. Where reports are produce with no purpose and no action.

Instead, the correct approach for implementing web analytics is to address all dimensions as depicted below:

It’s important to take baby steps in each and every area before growing too fast in one:

1. Governance – an organization should set clear ownership and guidelines for tracking, reporting, analysis, and action. The RACI model (Who’s responsible, accountable, consulted, and informed) can be used here to be crystal clear whose neck to choke when things go wrong.
2. Objectives – clearly define business objectives of every digital asset. For example, what’s the purpose of my website? Is it for leads generation or consumer engagement? Then define analytics metrics that reflect if my objectives are being met. In this case, if I’m after leads generation, then I might look at the conversion rate for media channels. However, if I’m after consumer engagement, then I’ll define measure goal achievement such as watch a video or click a share button.
3. Scope – too many companies want to tag everything when they just start with web analytics. Instead, be clear about what your objective is, and then set the scope to match each campaign/project. The best practice is to start with a single digital asset, whether it’d be a campaign site or brand site. Then expand to the entire digital ecosystem.
4. Team – this area is a pain point for many organizations. Too often, the web analytics role is filled by IT. Yes, IT can help in adding the tracking code. BUT, do they really know your digital marketing objectives? Hence, it’s important to have at least one single analyst dedicated to actually looking at the data and coming up with insights. This is THE SINGLE MOST important aspect of a successful analytics implementation.
5. Methodology – unfortunately, there isn’t a “best practice” way on what methodology you should use to setting tracking metrics and implementation. There’s only “your” way, because the methodology used should fit closely with the organization’s business objectives. Hence, tracking and measurement methodology should evolve close with the team’s experiences.
6. Tools – lastly, we can focus on identifying the right tool. From a core functionalities perspective, there aren’t that many striking difference amongst the web analytics tools out on the market. (e.g. GA, Baidu Tongji, Omniture, Webtrends, CoreMetrics, etc.) But if we look at from a long term data integration strategy perspective, there are some implications to data ownership and integration options.

So as you can see, web analytics is NOT just a technical tool. Doing it correctly can greatly enhance our digital marketing efforts. So I recommend everyone to carefully consider all the aforementioned factors in order to make an informed decision on web analytics implementation.

This article was written for ClickZ

The Purpose of Analytics

But before I dive any deeper, let me again ask a fundamental question: “What’s point of reporting and analytics within creative campaigns?” Well, most strategists/planners will say to measure the effectiveness of the particular campaign in driving strategic business objectives. And the creative/art directors will say to measure the effectiveness of tactical creative manifestations. With these two goals in mind, let’s look at some examples of creative agency reporting.

Mini/Brand Site Measurement Reports

One of the most common landing points for a digital campaign is the mini-site or brand site. Hence it’s only nature that we conduct in-depth reporting on this piece of digital assets. However, most reports produced by creative agencies serve none of the aforementioned purposes. Let me show you what I mean below:

Maybe the above report is enough for the judges at Cannes… BUT did it really give me any insights into why we’ve created this report in the first place? Keeping the two goals of analytics in mind, firstly knowing the bounce rate and UV/PV does NOTHING to help me understand if my digital asset is effectively serving my business objectives. Secondly, after looking at the top contents, I still have no clue how to optimize my creative to increase engagement.

Hence these reports are merely data puke, because we took the data from GA/Webtrends/Omniture and dumped it in a “Weekly Report” just for the hell of it. We don’t know why we’re doing it, we can’t get any business related insights from it, and most importantly I don’t know what ACTIONS to take afterwards. To better serve the purpose of analytics, let me suggested the following reports to measure digital campaign effectiveness:

From this report, I can gain insights on both aforementioned goals. The table on the left allow me to focus on metrics that tie directly with business objectives. As we look at the bigger picture metrics, we’ll begin to dig deeper and ask more tactical questions like “Why are my new registrations so low? Is it because of my content, and how can optimize my creative to increase my objectives?”

Those answers are delivered by the report on the right. The funnel analysis is one of the most effective ways to analyzing creative flow and user experience. From this report, we can see that in the 3 step process, 76% of users are not even making pass the 1^st step. This helps us narrow down user design issues as well as making more engaging content that drives the call to action.

Great! How Do I Get There?

As you probably gathered, the above report is not a one size fits all. As business objectives on digital vary greatly within industry and companies. So how do creative agencies work this out with their clients? Allow me propose the following model for implementation.

Digital Marketing and Measurement Model

This model is a structured approach for agencies to better tie digital assets with business objectives. It originated from the analytics expert Avinash Kaushik. A detailed step by step instruction can be found on his blog. Let me summarize the steps below:

1. Identify the business objectives for the digital channel
2. Set concrete goals to how the objectives manifests itself on the website
3. Selected a few key metrics as KPI’s as an indication for reaching each goal
4. Set a target for each KPI metrics based on industry benchmarks as well as previous campaigns
5. Identify the segments of people/behavior/outcomes to gauge effectiveness

An example of a completed model is shown below for a sample insurance client:

Conclusion

Using the above model, creative agencies can tie analytics to business objectives rather than reporting on shallow metrics to win creative awards. As I conclude this series, I’d like to emphasize the exact same point that I put forth in the previous column:

We as an industry need to understand that analytics is a tool to measure the delivery of digital marketing business objectives, rather than the verification of our own agency’s success metrics.

This article was written for ClickZ

With that said, let me show you some examples of typical agency reporting and what we can do to increase its effectiveness. Part 1 of this article will cover media agencies, and Part 2 will cover creative agencies.

Media Agency Post Buy

Since I work at a media agency, let me first point the dagger at our own industry by asking a fundamental question: “What’s point of reporting and analytics within digital media buying?”

Well, most media practitioners would say: “to measure the effectiveness and optimization of media buys.” BUT from my experience, we as an industry focus too much on “optimization”, and not enough on effectiveness. Let me show you what I mean by a sample report below:

Now what does this report tell me? From an optimization perspective, it indicates that this particular campaign (made up data) had lower CPC/CPM and a higher CTR comparing to the industry benchmark. Woohoo! Good job to the media planner, let’s celebrate! Wait… what about from a media effectiveness perspective? Impressions and clicks are high, so what? It tells me nothing about whether the traffic was effective in driving conversions. Instead, let me suggest the following report to show media buy effectiveness:

From this report I can see a breakdown of engagement, investment, and conversion by channel. Looking at a combination of Click vs. Cost, I can gauge the audience affinity by channel. An indication of an effective channel is: comparatively lower cost and higher clicks. But if we add the conversion metric to the equation, we can truly see how effective each channel is in driving conversions after the users clicked on the ad (display, video, social, etc.)

Great! Let’s Do It!

Hold on there tiger… while the above report seems easy to produce. There are actually a couple of core factors to why most of us don’t do it. These factors are: conversion complexity, and data silos.

1. Conversion Complexity

Let’s first talk about conversion point setting within a digital campaign. Campaigns with eCommerce are the easiest, with conversions being measured in transactions. However, most branding campaigns needs custom definitions of conversion, which may be: registrations, visits to the contact page, playing a game, etc. When you examine these activities, media buy is only one of the many factors that contribute to conversion. Things like user experience, content effectiveness, and design decisions are all factors can greatly increase/decrease conversion rate. Hence, it’s hard for a media agency to use conversion as KPI, because so many things are out of our hands.

2. Data Silos

This interesting problem lies in the fragmentation of tracking tools. Let’s explore this further by examining a typical campaign. Media agencies use their media tracking tools to track display ads on various media channels. When the user clicks on the ad, it directs them to a form of a digital asset: whether it’d be a mini-site, brand site, etc. At this point it enters the creative agencies realm, which use another set of web analytics tools to track user behavior and engagement. Now media data (impressions, clicks) and conversion data (user path, behavior) are completely segregated from each other in different databases. This makes it impossible to conduct cradle to the grave analysis from impressions all the way to conversion.

The Solution: Shared KPI’s & Data Centralization

The fundamental cause of the above issues really boils down to the lack of cooperation between media and creative agencies. As brands continue use a combination of different agencies (many of them in different media groups), the relationships amongst agencies are almost competitive, let alone cooperative. So in order to provide more effective analytics and reporting to our clients, we as an industry need to work closer together.

From measurement perspective, agencies should have shared KPI’s that measure the delivery of conversions, which ties directly to business results. Media agencies should no longer carry the attitude of “I delivered the impressions & traffic, whatever happens afterwards it’s not my problem”. And creative agencies needs to set clear measurement KPI’s to gauge creative and user experience effectiveness. (This will be covered more in part 2)

From a technology perspective, agencies should adopt the “single core platform” policy. Whether it’d be Omniture, Webtrends, or Google Analytics, make sure it’s used on both the display ad copies, as well as the digital assets. Of course, there will be more specialized tools used for specific purposes such as keyword optimization; these tools should be used in conjunction with the main tracking tool.

Conclusion

In the end, agencies need to understand that analytics is a tool to measure the delivery of digital marketing business objectives, rather than the verification of their own success metrics.

Next column I’ll be covering what creative agencies should do from their perspective to increase analytics and reporting effectiveness.

This article was written for ClickZ

The rise of digital shifted the various digital assets into marketing and agencies’ hands, where things could happen a lot faster and largely independent of IT.

However, today’s increasingly complex digital ecosystem calls for more and more technical initiatives such as: social CRM, digital analytics, and marketing intelligence data-warehouses. These initiatives have multiple touch points that are spread across both IT and marketing, hence the need for cooperation among marketers and IT.

Because IT has very different goals and operating models than marketing, this could cause frustration for marketers. As someone that moved from IT into marketing, let me share some do’s and don’ts for fellow marketers when working with the IT department.

Do
1. Do understand IT’s organizational objectives.
Newsflash! IT really doesn’t care about your marketing strategy. What keeps them up all night are things like system uptime, security, and data integrity. CIOs are usually measured by KPIs like: percentage of SLAs (service-level agreements) met, number of operational incidents, and number of security incidents. It’s important to note that all these KPIs are conservative measures to manage change.

So while you’re excited to put that new tool/dashboard in the company environment, IT is dreading how this change will impact their own department’s success metrics.

2. Do involve IT right from the get go.
Often, IT are brought in at the very last minute on marketing projects, usually right before deployment. This results in a whole array of issues that would ultimately delay or severely disrupt the go live process. The best thing to do is to have an IT member as a part of the core project team or steering committee. This ensures that IT is well informed and could prepare for any new software deployments.

3. Do conduct proper analysis and documentation.
When working with IT, especially on complex projects that involve data integration or workflow management, it is of prime importance to conduct proper requirements analysis and documentation. Do not run these projects like your usual Facebook campaigns! Most IT departments have a set development lifecycle that enforces requirements analysis so having these documentations ready would greatly enhance the working relationship with IT.

4. Do find out and follow IT’s security/compliance processes.
Most marketing departments are not aware of the complexities involved in deploying a new tool/system, especially within the internal environment. Every IT department has many processes in place to minimize risk. It’s important for marketing departments/agencies to understand processes and build in the proper time in the project plan to ensure compliance.

Don’ts
1. Don’t treat IT like solution providers.
Most internal IT teams consist of system admins, networking specialists, support, and maybe if you’re lucky some business analysts and project managers. But these folks do not have any proficiency in marketing technologies. Do not expect your IT to do web analytics (many companies still put this on IT), conduct SEO on your website, or do other seemingly “technical” things.

2. Don’t leave IT with maintenance and support.
The number one pet peeve for IT is to maintain someone else’s system/code. Yet they are left with maintenance when consultants and agencies leave after implementation. This is a huge risk for them because they are left accountable with codebase that they didn’t develop. Agencies should always build in maintenance and support as part of the service contract – IT will thank you for it!

3. Don’t give IT a deadline at the brief.
Back in my IT consulting days, I heard my fair share of “This is our project and I need it live in two weeks!” Most marketing departments will set a project deadline before even conducting requirements or analysis. This definitely does not sit well with IT as they have many concerns that need to be built into the project schedule. The best approach is to ask for lead time needed by IT for deployment and adjust the system to go live accordingly.

4. Don’t deviate too much from IT’s preferred SOE.
Each IT department usually has a preferred SOE (standard operating environment) for all software deployed internally. It’s helpful to understand the environment and try to seek technical solutions that fall within its boundaries. This would greatly speed up the deployment process as IT doesn’t have to spend too much time on auditing new technologies.

These tips are mainly for marketers to better understand the IT world. But it’s just as important for CIOs to understand the CMO’s challenges to increase cooperation between the two worlds,which would greatly enhance digital marketing effectiveness.

This article was written for ClickZ

Evolution of browsing behavior

After carefully reexamine my own sexuality, I have come to this conclusion: I am NOT gay for liking Pintrest. I am however, ridden with a bad case of ADD. But come to think about it, so is everyone else in today’s digital world. Don’t think you got ADD? Well, let me ask you the following questions:

1. When was the last time you read a book?… Completely
2. When was the last time you read a magazine? Sorry guys, Sports Illustrated Swim Suite Edition does not count!
3. When was the last time you read an online article? Skimming… sorta counts
4. When was the last time you read someone’s blog? Shit! Why the hell am I writing again?
5. When was the last time you read someone’s Tweet/FB Feed? … Probably within the last hour
6. When was the last time you skipped thru FB and looked at only the drunk/hot girls photos? … Probably 5 seconds ago

And there you have it… if you’re at least semi-digitally inclined you would notice that our information processing behavior have changed drastically with digital. Information are coming to us at lightning speeds, we can’t process everything, so we choose to process only the things that stimulate us. Which in this case… pictures. The guys at Pintrest are absolutely brilliant for noticing our browsing behavior evolution, and gave us exactly what we needed: visual stimulation.

How normal straight men (myself) use Pintrest

Anyways, now that the deep stuff is out the way. Let’s talk about some light-hearted stuff about how REAL men uses Pintrest!

First, and formost reason why I use Pintrest is… to look at pictures of pretty women, except I do it on the Chinese clone of Pintrest 拼图图. Why? um… because I’m an biological organism responding to my primal instincts to consistently seek out potential mates that can inherit my genetic specimen. One second thought, porn sites should really adapt Pintrest’s look and feel. I’m sure the time on site, return visits and overall “engagement” would be much higher.

Second reason why I use Pintrest is… to do spying/research/testing. Now I work in a digital agency, that means whenever ANYTHING come out with a bang on the internet. I gotta at least register an account and try it out. This includes my numerous social media profiles across almost all channels in both the U.S. and China, which I use a syncing service like 微博通 to keep up to date.

The third and final reason why I use Pintrest is… to scout for cute gifts for my girlfriend. Owwww, so sweet… bleh. Now, women are interesting creatures. I’d say most women’s buying behaviors are completely irrational and chaotic. They differ with men in that they buy base on “like” not on “need”. Hence the loads of outfits/shoes that they only wear once a freaking year while I buy shoes only when my last pair starts have holes in them. One interesting psychology that I’d like to point out is the concept of social proof. Women are innately more socially aware and conscious than men, hence they place great importance on being “popular” and “liked”. Now applying this concept on Pintrest, social re-pins means whatever the product portrayed is accepted by the masses, plop eCommerce on top that the product gets instant social proof. This is also the reason why many eCommerce sites in China are starting to follow the social eCommerce route adopting Pintrest’s model.

Okay, rant over… let me go back to browsing 美女 on 拼图图.

Renren was the earliest Facebook clone in China. It started as Xiaonei, which in Chinese means “On Campus”. Their primary user demographics are students, mostly college age in top tier cities. With their relatively young audience, the UGC (User Generated Content) is centered around light hearted content, as well as various other academic topic such as studying abroad, interviewing, and picking an major.

Kaxin001 is another Facebook clone with a heavy emphasis on social games. Their primary user demographics are urban white collars. Popular Content on the site includes several widespread and addictive mini-game such as Kaixin Garden (equivalent of Farmville on FB), and Park-It (mini game to park a car, of course with heavy product/brand placement). These social games caused quite a stir when first hitting the market, with dedicated players that get up at middle of the night in order to “Steal” vegetables from friend’s gardens.

Qzone is the biggest SNS by user base and traffic in China. Its primary demographics are teens and users from lower tiered cities. Qzone’s market dominance is largely driven by the integrated IM service QQ, which is the most popular IM tool in China. However, most users on QQ and Qzone do not use their real names. Instead they will most likely use an alias to identify themselves. So the content contains a lot more anonymity and is usually centered around the user’s daily lives. Also, Qzone is the only SNS that have a significant user base in the ages of 40-50. The reason is these parents all use QQ to communicate with their children, who are sometimes studying in another city.

Recognizing Qzone’s problem of anonymity, Tencent launched another real-name SNS service called Pengyou in 2008. Pengyou first started as QQ Xiaoyou, which in English means classmates. It’s primary demographics are again focused on teens and college aged students. But recognizing the narrowness of its focus, the name was officially changed to Pengyou (means friends in English) in the beginning of 2011. The growth strategy of this service is again to channel new users from QQ IM.

Douban is one of most unique SNS in the China market. The site has a large database of movies, books, and music, hence attracting a young crowd that’s heavily engaged in contemporary hip culture. The user demographics differ in social economic status, but are drawn because of common hobbies and interests. In terms of user base and traffic, it can’t compete with any of the bigger SNS’s. But it has one of the highest user engagement and loyalty. The content on the also differ quite a bit from the other SNS’s. The primary focus isn’t too much on the user themselves, instead heavily skewed towards their interests such as: movies they’ve seen, concerts they are going to, and book signing they’ve participated in.

51.com is also one of the earliest SNS’s in China. Its user-base age-wise is quiet varied, but are mostly from lower tiered cities. In terms of traffic and registered users, it closely follows Renren. But the social economic level and income is in no comparison with Kaixin001. The content on the site combines the mini-games from Kaixin and general SNS features like photo, blog, and status. Its recent direction is to shift to a social games platform that matches users gaming preferences, and provide social pairing suggestions.

In terms of market share, each SNS claim themselves to be the biggest. But due to the user segmentation, a lot of users will have multiple accounts on the different SNS’s in order to communicate with different groups of people. Also as the user-base matures they start to adopt a dominant SNS and abandon the others. From a SNS growth perspective, Tencent’s Pengyou and Qzone are ever rising due to QQ’s continuous stream of new users. Kaixin001 has been in decline due to the hugely popular Sina Weibo taking away its users. Renren is trying to expand outside of its young demographics but it’s been really an uphill battle.

In addressing the SNS’s market, we also can’t ignore the recent dominance of micro-blogs or Weibos as they are called in China. Eventhough the Weibos differ in feature from the traditional SNS’s, they are slowly becoming dominant form of SNS in China. Almost every big internet company in China has a Weibo service: Sohu, Sina, Tencent, 163NetEase… etc. But two dominant Weibo’s are Tencent and Sina. Feature-wise, these two services are almost identical. But again, consumer demographics are drastically different. Tencent Weibo gets most of its user from their QQ and Qzone services, hence the demographics of its users are mostly teens. While Sina Weibo focus more on the white collar demographics. Content-wise, Tencent is mostly daily life focused, while Sina has lot of professional and industry content. So users of Sina would find the content on Tencent absolutely childish, and Tencent users would find the content on Sina very boring.

This article was written for ValueCap Research