Read More

Want to review a new digital camera, get gift ideas, or buy tickets to the next Morrissey concert? If you're in Indonesia, KASKUS is your place. 28 million unique users buy, sell, talk and share information on the site each month, making it the country's largest user-generated content publisher.  

With so many users, KASKUS recently faced a growing challenge: how to serve ads that are relevant to users’ age, gender and interests? 

“As KASKUS is the leading digital community and social commerce platform, our vision is to drive data-driven monetization by making our first-party audience data actionable, we want to give advertisers ways to perform better in our sites and increase the effectiveness of our impression-based ads." Ronny W. Sugiadha, Chief Marketing Officer for KASKUS

Sugiadha and his team wanted to create an audience segment that had a high demand among advertisers: users who had shown interest in mobile devices and were more likely to purchase them. 

KASKUS turned to Sparkline, a Google Analytics 360 Services and Sales Partner, who worked with them to approach the challenge to serve the most relevant ads. The process went from an advanced Google Analytics 360 implementation, to segmentation analysis and audience sharing with Doubleclick for Publishers (DFP). 

Below is a screenshot of the actual segment shared between Google Analytics 360 and DFP. To learn more about the process, read the full case study. 

How well did the new audience work compared to its old open-auction inventory in the Doubleclick Ad Exchange (AdX)? 

"Using the Google Analytics 360 Audience Segment sharing feature in DFP and AdX, we doubled our CTR and saw a 3.3X CPM uplift on this audience-targeted AdX inventory," reports Ronny Sugiadha. "We are looking forward to even more positive impact moving forward."

To learn more about how KASKUS achieved those results read the full case study. 

Posted by Catherine Candano and Daniel Waisberg, Google Analytics team

Read More

Hello, I'm Mark Edmondson and I have the honour of being a Google Developer Expert for Google Analytics, a role that looks to help developers get the most out of Google Analytics. My specialities include Google APIs and data programming, which has prompted the creation of googleAnalyticsR, a new R package to interact with the recently released Google Analytics Reporting API V4.

R is increasingly popular with web analysts due to its powerful data processing, statistics and visualisation capabilities. A large part of R’s strength in data analysis comes from its ever increasing range of open source packages. googleAnalyticsR allows you to download your Google Analytics data straight into an R session, which you could then use with other R packages to create insight and action from your data.

As well as v3 API capabilities, googleAnalyticsR also includes features unique to v4:

•  On the fly calculated metrics 
• Pivot reports 
• Histogram data 
• Multiple and more advanced segments 
• Multi-date requests 
• Cohorts 
• Batched reports 

The library will also take advantage of any new aspects of the V4 API as it develops.

Getting started

To start using googleAnalyticsR, make sure you have the latest versions of R and (optionally) the R IDE, RStudio

Start up RStudio, and install the package via:

install.packages("googleAnalyticsR")

This will install the package on your computer plus any dependencies.

After successful installation, you can load the library via library(googleAnalyticsR), and read the documentation within R via ?googleAnalyticsR, or on the package website.

An example API call — calculated metrics

Once installed, you can get your Google Analytics data similarly to the example below, which fetches an on-the-fly calculated metric:

library(googleAnalyticsR)

# authenticate with your Google Analytics login
ga_auth()

# call google analytics v4
ga4 <- google_analytics_4(viewId = 123456,
                         date_range = c("2016-01-01",
                                       "2016-06-01"),
                         metrics = c(calc1='ga:sessions /
                                            ga:users'),
                         dimensions = 'medium')

See more examples on the v4 help page.

Segment Builder RStudio Addin

One of the powerful new features of the v4 API is enhanced segmentation, however they can be complicated to configure. To help with this, an RStudio Addin has been added which gives you a UI within RStudio to configure the segment object. To use, install the library in RStudio then select the segment builder from the Addin menu. 

Create your own Google Analytics 

Dashboards googleAnalyticsR has been built to be compatible with Shiny, a web application framework for R.  It includes functions to make Google Analytics dashboards as easy as possible, along with login functions for your end users. 

Example code for you to create your own Shiny dashboards is on the website.

BigQuery Google Analytics 360 exports 

In addition to the v4 and v3 API functions, BigQuery exports from Google Analytics 360 can also be directly queried, letting you download millions of rows of unsampled data.

Aimed at analysts familiar with Google Analytics but not SQL, it creates the SQL for you to query common standard metrics and dimensions, using a similar interface as the API calls.  See the BigQuery section on the website for more details.

Anti-sampling 

To more easily fetch non-sampled data, googleAnalyticsR also features an anti-sampling flag which splits the API calls into self-adjusting time windows that are under the session sampling limit.  The approach used is described in more detail here.

Get involved 

If you have any suggestions, bug reports or have any ideas you would like to contribute, then you are very welcome to raise an issue or submit a pull request at the googleAnalyticsR Github repository, or ping me on Twitter at @HoloMarkeD.

Posted by Mark Edmondson, Google Developer Expert

Read More

Your organization has plenty of data about customer behavior that tells you what different customers do where and when. You can see when they visit you online, how long they search, and how much they spend.

But too often the “why" behind their actions remains elusive. With the mountains of information you collect, the insights are often difficult to find, take too much time to discern, or require additional data. All this means it takes marketers too long to get important information that could make a real difference to the customer experience — and the bottom line.

“If you want to have a major impact, you need an integrated approach to see what is happening at all customer touch points and understand how effective you are,” says Joerg Niessing, a marketing professor at INSEAD.

The number of sources of marketing and customer data that a company integrates correlates strongly to performance vis-à-vis competitors, according to a recent study published by INSEAD. The study focused on customer and marketing data, including:

• Digital analytics, such as optimizing email campaigns, testing content, and analyzing digital pathways to optimize website use and experience.
• Customer analytics, including lifetime value and loyalty calculations, response and purchase propensity modeling, and micro segmentation.
• Marketing analytics, such as demand forecasting, marketing attribution models, market mix modeling, and media budget optimization.
• Sales analytics, including pricing elasticity modeling, assortment planning, and sales territory design.
• Consumer analytics, including surveys/questionnaires, customer experience research, and customer satisfaction/advocacy modeling.

The study found that those companies that leverage multiple sources and focus diligently on demand generation have significantly stronger business performance, especially total shareholder return.

Straight to the source

But insights uncovered from many data sources often beg the question, “Why?” To answer that, modern marketers go directly to the source: consumers.

Traditionally, companies that use surveys and field research to try to get at the “why” behind the “what” pay a lot of money for information that is often too complex to understand and too slow to arrive. When it does come in, it is sometimes no longer relevant and leaves organizations trying to solve last month’s or last year’s problem at the expense of current ones. Attempting to get speedier or less costly results risks compromising accuracy.

But innovations in market research are changing the game. Easy-to-use survey tools like Google Surveys help marketers fill out their knowledge of customer behavior much faster than traditional surveying methods.

Companies that make use of these fast, convenient survey solutions gain insight not only into what people actually do, but also what they say they will do — and in that gap there could be opportunities. “Marrying digital and marketing analytics with consumer research from surveys gives marketers deeper insights and opens up the number of hypotheses a company can test,” says Suzanne Mumford, Head of Marketing for the Google Analytics 360 Suite. “Marketing today is in near real time and your data should be, too.”

“Marrying digital and marketing analytics with consumer research from surveys gives marketers deeper insights and opens up the number of hypotheses a company can test.”
—Suzanne Mumford, Head of Marketing, Google Analytics 360 Suite

Say your website analytics reveal that one segment of your visitors are highly engaged with your site content, but their visits aren’t converting into sales. “You can ask them directly why they keep coming back but don’t end up buying. Surveys let you take your data one step further and round out the picture of the customer so you can make informed business decisions and tailor your customer experiences,” says Kevin Fields, Product Marketing Manager for Google Surveys.

Supporting business decisions with surveys

Surveys are also useful if marketers find themselves in an internal debate about two campaign concepts. Before making a large investment based on subjective opinion, marketing leaders can validate messaging by asking the target audience about their preference.

For modern marketers, surveys have become an essential element in an integrated marketing approach — they produce insights that complement those uncovered by other data sources. “I want to make sure that the customer voice is front and center but that we also surround it with other data — that we can make really good, holistic business decisions,” says Stacey Symonds, Senior Director for Consumer Insights at Orbitz.

So think about what you’d most like to ask your customers — or those who visit your site but don’t buy. Survey solutions like Google Surveys allow businesses to get sophisticated, accurate data in a matter of days, not months. Because these methods are more affordable and quick, they allow businesses to continually iterate to meet customers’ needs.

“Surveys empower organizations to get answers when they matter,” Fields says. “And getting those insights quickly helps marketing stay nimble.”

Download “Measuring Marketing Insights,” an online Insight Center Collection of articles from Harvard Business Review, to learn how organizations are using market research to gain more consumer insights.

A version of this article first appeared as sponsor content on HBR.org in August 2016.

Posted by Kevin Fields, Product Marketing Manager, Google Surveys team

Read More

One of the biggest challenges for marketing leaders today is not finding or hiring analytic talent, according to new research cited in a Harvard Business Review report, but rather it is finding the right ways to move the mountains of data into insights and then into action.

The study concluded that marketing organizations need analytics professionals who understand data and the technologies that collect, house, and integrate it.¹ That’s a given. But beyond that, experts say, executives need to place more emphasis on data science than on data scientists. Put another way: They should pay more attention to analyzing and acting on what they have now because analysis paralysis doesn’t create customer value.

“Data scientists are technicians who are very good at managing and manipulating data,” says Peter Fader, the Frances and Pei-Yuan Chia Professor of Marketing at the Wharton School of the University of Pennsylvania and author of Customer Centricity: Focus on the Right Customers for Strategic Advantage. “But data science is about looking for patterns, coming up with hypotheses, testing them, and acting on the results.”

Machine Learning

That’s where machine learning can speed analysis and augment your analytics team’s work — by crunching massive amounts of data to identify patterns and anomalies.

A type of artificial intelligence that uses algorithms that iteratively learn from data, machine learning can surface insights without being explicitly programmed where to look for them. It makes it more efficient to crunch massive amounts of data, calling out issues before you see them and providing answers to questions you may not have even thought to ask. This speed to insight allows marketers and analysts to do more with the data that comes in and see the whole picture of the customer journey.

Accenture Managing Partner Conor McGovern says, “If you can’t make the rubber hit the road with a disciplined approach to analytics, you will end up with customer experiences that aren’t as effective or engaging as they could be. As with any source of information, you need to embed and ingrain analytics into decision-making processes to obtain the desired results.”

“If you can’t make the rubber hit the road with a disciplined approach to analytics, you will end up with customer experiences that aren’t as effective or engaging as they could be.” —Conor McGovern, Managing Partner, Accenture

How Lenovo Harnessed Data to Create Customer Value

That targeted data science approach can give companies of any size a competitive advantage. Lenovo is a prime example of a marketing team that mastered the use of advanced technology and analytics tools, driving the company to create better value for its customers.

Ajit Sivadasan, Vice President and General Manager of Global E-commerce, realized that customer data was burgeoning and Lenovo needed to harness it. He began by establishing an analytics team in his e-commerce unit that today integrates and analyzes customer and marketing data from more than 60 sources worldwide. By integrating and analyzing Lenovo’s data, Sivadasan found that there are three main drivers of customer satisfaction that correlate to loyalty:

1. Quality of the online experience. Sivadasan’s team tracks important variables such as how easy it is to find product information and whether Lenovo provides sufficient follow-up on the status of an order.
2. Meeting commitments. This second driver includes how often the company misses promised ship dates.
3. Experience with the product itself. By analyzing social media and direct customer feedback, Lenovo’s ecommerce team helps the company improve its products.

Competing on Analytics

In order to pursue an effective analytics strategy, executives have to clearly define business problems and what the questions are that analytics can answer. If executives don’t do this, they risk getting back data that sends the organization in the wrong direction.

For example, companies frequently find themselves puzzling over a dip in conversions among a desired demographic. Organizations need to be able to study the data, ask customers and potential customers the right questions, and experiment with offering different solutions to optimize the customer experience. Answers need to come in quickly so the organization can act quickly — ahead of the competition.

The speed to insight that machine learning offers can help companies act strategically on the data they have, homing in on the insights with impact, allowing executives to make informed decisions.

Says Joerg Niessing, Marketing Professor at INSEAD: “Executives still have to make the same strategic decisions that they have always made. They need to understand market dynamics and what competitors are doing — and then determine how the company should react. The only difference is that we now have a great deal more data and analytics to help make these decisions.”

Download “Measuring Marketing Insights: Turning Data Into Action,” an online Insight Center Collection of articles from Harvard Business Review, to learn more about using analytics to create customer value.

A version of this article first appeared as sponsor content on HBR.org in August 2016.

¹Harvard Business Review Analytic Services, "Marketing in the Driver's Seat: Using Analytics to Create Customer Value," 2015.

Posted by Karen Budell, Content Marketing Manager, Google Analytics 360 Suite

Read More

Almost every organization today is putting customer experience (CX) at the core of its strategy, aiming to provide products and services that meet customers at every touch point. In a crowded, multichannel marketplace, companies realize that a great customer experience — consistently delivering what customers want, when they want it — can be a powerful differentiator.

But many companies fail to deliver, according to research by Harvard Business Review Analytic Services (HBR-AS). Although half of surveyed business leaders say CX is a top-two differentiator for their business, just half of them said they perform well in it.

Although half of surveyed business leaders say CX is a top-two differentiator for their business, just half of them said they perform well in it.¹

The problem isn’t access to data; most businesses said they collect mountains of information on their customers. The real obstacle to better customer experience, the research has found, is built into the way organizations share that data, analyze it, and work together.

Improving the customer experience is the end game, but getting there requires more than data. It requires the right data, from multiple channels, integrated to give a holistic picture of the customer journey. And that is where many companies struggle. HBR-AS found that fewer than a quarter of companies integrate customer data across channels to provide a single customer view.

Integrating data for customer value requires getting around organizational silos, which HBR-AS research has identified as the number one problem for companies struggling to improve their total customer experience. These silos prevent organizations from understanding the customers’ expectations at critical moments, and cultural resistance makes it tough to get the collaboration needed to solve the problem. As a result, respondents said the business doesn’t develop the right insights, get the information to the right people, or make the moves that could add real value.

Data-Driven Insight

By contrast, the study found that “best-in-class companies” — those with strong financial performance and competitive customer experiences — are more likely to have broken down those silos than are other organizations. And they use sophisticated analytics in a way that provides insights that open up the customer experience to the whole organization.

For example, at Progressive Insurance, the marketing team collected data on how mobile app users were behaving. These consumers, they discovered, wanted more than just helpful insurance quotes in the mobile app; they wanted to buy insurance on the spot. Progressive responded by giving them exactly what they wanted — the option to buy insurance — which vastly improved the customer experience and delivered a big win for the company. When a company creates customer value, the business benefits naturally follow.

Marketing Takes the Lead

But who is going to break down silos, connect the dots of the customer experience, and drive its improvement?

Today, marketing leaders need to make the case to the company that optimizing the customer experience requires breaking down silos and opening up collaboration, and shifting from a product-centric to a customer-centric approach, says Erich Joachimsthaler, author of Brand Leadership: Building Assets in an Information Economy. For example, a European beverage company assigns marketing groups to consumption moments, such as a night out, instead of brands and channels. The goal is to embed marketers deeply into a particular customer experience and focus them on each step of the customer journey.

“Marketing needs to connect the dots across all customer-facing functions of a company, including partners, in order to deliver real value instead of just communicating the brand,” says Joachimsthaler.

Robust analytics and insights have given marketing teams insight into how customers interact with brands, highlighting product preferences, purchase sequences, and so forth. And they reveal how top of the funnel marketing activities — such as an online display ad or TV commercial — tie in to in-store sales or an online website conversion. Measurement and analytics allow brand marketing and performance marketing to complement each other for the customers’ benefit.

Clearly the stakes are high, and marketing leaders and their teams are challenged to think in new ways. They don’t need more data; they need to find ways to identify and supply their organization with useful insights from that data.

Download “Measuring Marketing Insights,” a collection of Harvard Business Review Insight Center articles, to learn how companies are using data and marketing analytics to improve customer experience.

A version of this article first appeared as sponsor content on HBR.org in August 2016.

¹Source: Harvard Business Review Analytic Services, "Marketing in the Driver's Seat: Using Analytics to Create Customer Value," 2015.

Posted by Karen Budell, Content Marketing Manager, Google Analytics 360 Suite