09:03 am, idea-stack
1 note

05:34 pm, idea-stack
picture HD
Pinteresting’ Facts About Pinterest Users [INFOGRAPHIC]

Pinteresting’ Facts About Pinterest Users [INFOGRAPHIC]

04:49 pm, idea-stack
1 note

12:11 pm, idea-stack
1 note
Advanced Multi-Channel Funnel Analysis using Google Analytics

Multi-Channel Funnel has been in Google Analytics for a while. Although by searching  "Multi-Channel Analysis"  you could find a lot of great how-to articles to leverage this powerful function, but seldom of them have explored the opportunities in using it for better resources allocation decision based on ROI estimation, particularly, using Assisted Conversions. Hence, i have decided to put together my experience in marketing, analytic, and infographic to demonstrate the following analytic model. Enjoy.

(Disclosure: i am currently working at MRM Worldiwide, a Digital Strategy agency under McCann WorldGroup, and hopefully the following model will be used in our service someday…..so…….. have fun ! XD)

A Closer Look to Assisted Conversion

Apart from the Multi-Channel Funnel view in Google Analytics, Assisted Conversion Report is the one that we are looking for. If you have some experience in Omniture, you would know that Assisted Conversion is indeed having similar logic as Participation Variable, a way to estimate the potential value that a particular entity, exists within a funnel of process, has driven. For example, if a visitor purchase a dress online after she visit a review on a forum (outside the eshop), both our eshop and the external forum will be entitled to have contributions, in terms of conversions and revenue, counted towards themselves, either evenly (all entities have the same value) or linearly (all entities gain the average of value gained, only in Omniture), demonstrating that how the entity “participate” within the whole conversion path.

A snapshot of Assisted Conversion Report of my groupbuying site: Cheapppy

What makes Multi-Channel Funnel in Google Analytics more powerful (than Omniture) is, it has segmented that participation value for you, based on whether that entity, in the above graph would be “Channel”, has contributed as the Last Interaction or not. Should that channel is not the last step before the visitor being converted, then it’s “assisting” the conversion flow, and that counts towards as their Assisted Conversions instead.

The power behind this logic is the different between Assisted Conversion and Last Interactions Conversion. Traditionally we count conversion towards the “last stop” of visitors, but with the uprising of Social Media, where fans usually “engage” and “consider” rather than “converted”, increases the complexity of the tradition conversion path as well as the evaluation process. Google does this tedious work for you, by introducing the Assisted Conversion, it is easier for analyst to tell if certain channels are good enough to support the conversion flow despite that they might not be the “last stop” of visitors. To make this concept even more clear, Google introduce the “Assisted / Last Interaction Conversions (Ratio)” which tells whether a channel could drive more Assisted or Last Interaction Conversions (>1 = “contribute more within the flow”, <1 = “contribute more as last stop”).

So, How to decide when we need Resources Reallocation?

Before drilling into details on how we could leverage such report for resources planning, let’s talk more about how to determine a if a channel is “Good-or-Bad” under the new complexity of conversion cycle.

Knowing the Path is one thing, determine the effectiveness is another
To answer such question, indeed, it depends on the “what you are looking for”. In general, a channel with high ROI (relative to other channels) would always mean that they’re performing better. With the help of Google for having segmentation in Assisted between Last Interaction makes this question more insightful: is certain channel better at assisting other channels for conversion? If i were Levis, should my f-commerce strategy more effective in assisting other channels or driving direct conversion?

Another frequently asked questions would be, instead of Channel level, how good would certain Ad Group performing? How good are we adapting our Sales cycle along with our SEM strategy (i.e. paid search traffics driven by targeting relevant landing pages based on Awareness-Consideration-PurchaseIntent model) ? An effective Ad Group should thus have a relatively higher Assisted ROI if it is targeting for Awareness or Consideration items, otherwise we should change the target to Purchase Intent or even Conversion pages if it’s last interaction ROI is higher.

Simply speaking, to determine if we reallocating from one to another, we need to determine the characteristics of the items, either they’re Channels or AdGroups, first.

Assisted ROI & Direct ROI Estimation

Normally Google suggests us to us “Assisted / Last Interaction Conversions” to analyze for how good certain channel perform in either Assisted or Direct way. But it’s just a ratio based on “occurrence”. If you have followed the whole logic so far, you should know by now we should look for a way to determine both the Assisted ROI (from Assisted Conversions) or Direct ROI (from Last Interaction Conversions) of Channels which provide a more business angle for us to handle our question. So how could we calculate such business metric based on what we currently have (the Assisted Conversion report)? Let’s begin from the basic definition:

As for our case, we could easily fill-in-blank using the following formulas…

And here’s how the data (fake one) presented in a spreadsheet:

Spreadsheet with mock-up data (already explained a lot of things!)

Organ Part - Assisted Conversions Report from Google Analytics 
Green Part - Cost spent a particular Channel (or items). Based on the participation concept, the total Cost spent on a Channel will be shared by all the Assisted and Direct Conversions, thus the Assisted Cost and Direct Cost of a Channel could be estimated by the portion of corresponding conversions achieved.
Purple Part - ROI based on corresponding segment (Assisted or Direct)

Easy enough, to determine the channel characteristics, we simply put the data on a Relational Map, with x-axis as Assisted ROI and y-axis as Direct ROI. Almost done!

Looks like the Referral is under-preformed, sounds like a good action point to begin with!

What Actions We need to Take?

I couldn’t emphasize more enough that any chart or infographics without Action Triggers is simply meaningless. In my last post, where I have talked about how we could put an infographic onto an upper level and make it more actionable in the 5th steps. A relation map like above is no more than a graphics presentation based on data. So the key is to help readers identify the action right away after they read the chart. On the above chart, think about breaking down into 4 different sections:

1 | 2
3 | 4
Section #1 - Channels that are bad in Assisted Conversion (-ve Assisted ROI) but good at Direct Conversion (+ Direct ROI)  
Section #2 - Channels that are good at both Assisted and Direct Conversion 
Section #3 - Channels that are bad at both Assisted and Direct Conversion…. (simply under-preformed…) 
Section #4 - Channels that are good at Assisted Conversion but bad at Direct one 

Now identifying action is simple:

If you are looking for under-performed channel, read Section #3, those channels are definitely having some issue (which you have to investigate!)

If you are looking for performance of social entities, like Facebook, Twitter, Youtube, and even other online communities,  where you are expecting high Assisted ROI, see if they’re at Section #2 or #4, if not, well, you know what i mean.

How about reallocation of resources? Start from something poorly performed and move those resources, usually dollar-signed, to those well performed ones. Simply speaking, if you have decide to reallocate instead of optimizing or fixing problem, move resources from items in Section #3 to those in #2. Easy.

The point here is, in any business, or even down to a single business process (like SEM), all we’re focusing is Return on Investment. If some operation couldn’t bring return in any form (e.g. Impression is a kind of return, and we could easy convert it into dollar-signed using CPM), then it is either we need to fix the problem, or simply give it up and free the resources for those well-performed ones.

Always remember the if-this-than-that rules. It always helps in designing for actions triggers.

Looks Good, but only Channel level Analysis?

No. (Why stop here? XD)

The Assisted vs Direct ROI drives lots of potential dimensions in analysis, here’s some other variation we could take a look based on the same logic as above:

1. Referral Analysis
Segment the Assisted report based on Source/Medium will give you a over view in all upstream traffics. It is essentially important if you have broad social media strategy which occupying different channels like facebook plus twitter plus linkedin plus pinterest and so on… then such break down will let you understand how good is your social media team is working and let them know if they need to tune their tactics in different media.

2. Campaign Analysis
A more aggressive approach is broken down by Campaigns. This angle provides marketers a more insightful view on how different campaigns are performing. Not limited to social media engagement, but also social ad. vs sponsored tweets, break-up email vs cart-abandon emails, banners ad vs display ad, (offline with qr code) etc. Make sure your marketing team have tagged the upstream URL correctly in order to fully leverage this powerful report.

3. AdGroup / Keywords Analysis (in AdWords)
The last angle we could have a look is the SEM performance, which, traditionally, we focus too much solely on the click-through rate and cost per click, and simply overlook the importance of how they actually generate goals or even leads to our business. With the Assisted / Direct ROI model we could now easily tell if certain AdGroup or Keyword are performing as expected, says, a set of retention keywords (e.g. “where to repair my iPad 2?”) should be expecting a higher Assisted ROI as it helps satisfying customer and supporting future purchases. This will also help strategizing how each landing page should be doing as well.

——————————-(i’m just a <hr/> tag……)—————————————

I guess that’s for the looooong class.  (my bad style….. XD)

How much are you convinced by the model? Have you tried feeding in your real business data and see how they work out? Drop me a line if you have any comments or questions ! Would love to hear from you !

Again, subscribe me if you haven’t,  if you like, and feel free to connect with me, too !

Coming soon will be the “Framework for Business Infographics Design - v1.0”, so Stay Tuned !!!~~


Other Stories you may be interested:

6 Steps to Crack a Complex Business Report into Actionable Infographics 
[How to] Crack a Complex Business Report into Actionable Infographics (Part 1) [Long Story Alert !] 
4 Things Startups Should Learn from Moneyball

11:58 am, idea-stack
4 Things Startups Should Learn from Moneyball

Finally have some free time to go through this masterpiece. Just as the poster on the right said: “It’s more than a movie…”. Despite it talked mainly about how GM Billy fight the long way for his belief in statistical analysis for baseball, indeed, we have so much to learn from this old story in 2002…. particularly, startups.

[ kindly warning: spoilers alert. If you haven’t watched this movie yet, skip to the end, quickly! :) ]

1. It’s All about Talents

At the beginning of the movie, Billy quickly recognized Peter Brand, a fictional character based on Paul DePodesta, as Peter has been influencing the GM of Cleverland Indians in making decision even though Peter is just a graduate. Billy then talked with this young boy regarding his enlightening insights in analytical approach in baseball: “Buy wins, buy runs, but not players”. While by that time Billy was planned to trade players with Cleverland Indians, but he turned out bought Peter, a player analyst instead. Throughout the rest of the movie, Peter proves his valuable supports to Billy and brings Oakland to the winning roads. All these won’t happen if Billy haven’t treasured Peter’s ideas, and most importantly, treasured Peter himself as a special talent.

2. Always be Faithful to your Belief

Picking the right talents is one thing, treasuring them and have faith on them is another. But don’t mis-understand this, Billy put his belief on Peter is not because of personal matters (or just a little), but Billy himself does have faith in that “statistical system”, and even more he does hold his faith almost blindly even though when the whole scouting team plus the team manager Art Howe were all against him. But hey, if you think that it works then why bother how the other says? Billy then fire the scout lead and trade all the traditional players to force Howe to play as Billy wants him to. Yes this won’t please everyone, and it never does, yet if you do believe it will be successful, stay on your track, knock them off your way and keep moving. Please them with your success at the end and thanks them to motivate you negatively all the way long.

3. Have Balance between Intuitive and Analytics

People usually believes that startups are all about guts feeling and intuitive on certain visions, and often, they against analysis, just because they think that it’s an art of big enterprises and should never be used in startups as a key role or even as a culture. Billy in the firm dose drive the whole movie based on his guts feeling on Peter’s theory initially, but fundamentally, he is indeed convinced by how the stories that could be told within those data, and decide to pick under-value players solely based on his KPI, the On-Base Percentage (OBP), as well as other sabermetrics, as the key of his later scouting strategy, which, has been proven to be correct. In fact, some startups nowadays have already embracing lean analytic approach for quick pivoting and MVP (Minimum Viable Product). Some of them, like Buffer, which used a set of three-pages as MVP to test for customers validation based on click through and turns out to launch their apps in 7 weeks from idea with paying customers and funding from Guy Kawasaki, simply proves that startups do need objective analytic for better survivorship in this bloody red-sea competition on innovation. Don’t say that gathering data is slow and sample size is too small for analysis, it’s only depends on how you leverage and tell stories from those data.

4. Keep Evolving and Staying on the Edge

The movie ends by 2002, after Oakland Athletics wins their 20-in-a-row match, and Peter rejected the offers from Red Fox as the highest salary general manager, however, the story hasn’t ended yet. By 2002, the book Moneyball (ISBN 0-393-05765-8), which the movie is referring to, has been published because of Billy’s stunning success on his strategy, and this has made quite a lot of baseball team by that time started using his system to optimize the ROI of their team, and of course, casing Athletics starts losing their edge. But in reality, Billy hasn’t given up his belief, instead, kept evolving it. After a few years of low winning %, he slowly shifted their scouting strategy to other under-value players based on defensive skills and even choosing high-school players, which he considered as usually-undervalued ones, to, again, optimize the winning chance of Athletics. As it turns out, by emphasizing defense they managed to reach back 50% winning chance. Although they still miss the playoffs, but their strategy is evolving, and always trying to stay on the edge with their style. Billy is continuing his new contract till 2019, and this we can obvious see that his legend keeps going on, and on.

Now it’s your turn, what do you think about the firm? Have you learnt anything from it (apart from Brad Pitt is sooooo cool)? Share with us in the comment!

And don’t forget to subscribe my blog or connect with me if you wish to !

See you next time.


04:24 pm, idea-stack
6 Steps to Crack a Complex Business Report into Actionable Infographics


After the release of Part 1 and 2 I’ve so much positive feedback from friends, and the most asked question from them is “When the Part 3 will be released?”… Well, sorry to say, as start getting busy because of a huge analytics project from a key account, plus as in my draft, the last part would involve relatively more images to demonstrate the “Action Triggers” design, so i guess it would take a few more weeks to bake the stuff. But i guess we have no point to cool down and stop learning about this field, right? So let’s have a short step-by-step guides as an overview before the last part!

6 steps, are really all we need to make an actionable infographic for analytic.


1. Start with something *very* painful

Reports usually act as an essential catalyst within chain of business process for decision making. While in the business class we have learnt that accurate and detail data is the key to make correct decision with precision, yet in reality, the reports usually end up with huge matrix of metrics which is simply too chaotic for human eyes. Some better ones would have bars charts or trends lines to reflect the data in different dimension, yet all those charts could do is, well, really reflecting the fact, they still lack of a true purpose that would like to help the reader to achieve. In short, they’re simply visually and mentally painful for a business person to decide the right decision. And this suggest a very good starting point for our cracking process.

2. Learn no Objective but how Users Read and Think

When we would like crack something, the first thing we have to do is to try approaching the same problem in different angle. If we think of how the report was designed in the first place, we would always go back to the origin, a list of so-called “Objectives” of the report. I am not saying that those objective are not useful, and clearly they were once asked by seniors who hope these report could deal with certain business problem, it’s just our old knowledge from business school turn these questions into something visually more complex questions….

How could we deal with this paradox? Think differently. If a report was designed with all objectives met and have already been using for long (says a year) within a process chain (e.g. deciding how to allocate our marketing budget), then we need not to challenge the importance of the report, as what it need to be optimized is its effectiveness. Think a step further, if seniors have been using this one for decision making for some long already, then it means that that painful report is “really telling something”. So what we need to do is indeed simple: learn how the users now read and think. Why? Because after the painful day they have already learn what area they should read in their first glance, and how to combine different visual or numeric difference to draw a conclusion! So ask in details from the order and the scope of metrics that they look at (says they usually only interested in last 3 months rather than last year), to how they pick up different visual effect (like a uprising trend or a green cell because of positive changes), and finally reaching a conclusion. Now this is your Use Case which tell you what component you need for your designs to be working for them and you are ready to start designing.

3. Nailing down the Metrics and Design & 4. Time to Draw Lines, Circles and Squares

It’s always a good idea to go these two steps hand-in-hand, such that while you’re thinking about the metrics, you will also design for the graphics and which will provide you visual clue that helps you to refine the metric afterwards (e.g. I have last 3 months of sales data > but i need only one abstraction metric for the x-axis on Map chart > so i use average based on last 3 months instead).

When thinking about metrics, consider the scope and some other visual clues that you have gather, you should have much you need to proceed this step. And believe it or not, this step requires a lot of analytic experience to support accurate decision in choosing, customizing and nailing down the right metrics and design for the next coming steps. Particularly, if the user was depending on, says, 5 different on-site engagement metrics to determine the content consumption effectiveness of a single page, how could we simplified the overwhelming details (remember, it’s all about assisting in right and insightful decision making but not the details of data!) ? Could we create a custom metrics like “Engagement Index” for each page by massaging (like, multiplying) those 5 different parameters? Yeah, analytic knowledge helps here.

When comes the drawing part. While you might think that “well i am not a designer…”, i would say you are wrong. Stick to the basic visual components: Lines for dimensions (page view, engagement index, time, etc.) , Squares and Circles for items with size and colors for any two additional metrics, and as long as the users questions require them to make decision by comparing different object (i would say 70% business problem falls in this group…) use a Map charts. Always remember a Map chart provides the greatest flexibility for supporting multiple variables in a tiny yet informative 2D space. When in leisure, learn more how other designs their infographic on visual.ly and visual.org, and take some time to read about the Napkin from Dan Roam and also his new book Blah Blah Blah, and of course never miss the bible, The Visual Display of Quantitative Information from Edward R. Tufte,  all these are great readings would definitely boost your sense in designing infographics.

In any case, a quick tips for you would be: if you end up with 1 metrics, think a number or a bar; 2 metrics and one is continuous (like time) then think time series charts, otherwise for discrete data, think scatter map; 3-5 metrics, think a Map with customized items (like circle with size tie to one metrics); and for 6 or above, please re-do the step 3 again or considering break the graph into two…

5. Design Actions Triggers with Arrows

Despite going that far, one thing i would like to stress again is the final design will be embedded back into the business process chain, which means, either it is very simple enough and no need to be explained for decision, or it is very well defined with what i called Action Triggers. Defining an AT is indeed a if-this-then-that rules for the designed chart. Says for example, if the length of this metric bar is taller than this line (a threshold drawn along with a bar), then i have to do this and do that. Another example would be if certain item falls into the bottom-left section (of a Map), then i will pick one item in upper-right section and do something for them. Simply speaking, identify rules for user to interact with the infographics, and use arrows to indicate the detail interaction (like in the second example, direct user from bottom-left to upper-right with an arrows). As long as all triggers are listed out, we can easily define the corresponding actions we have to take for this particular observation. The key here is streamlining from how the user read the graph to making decision. Sure we could introduce a lot more different kind of actions to be triggered, like for some situation we won’t take action but takes note on it and put it onto a monitoring stack for next time decision. This might sound a little bit difficult to follow, but i hope you get the concept here. And i will cover this more in the coming Part 3.

6. Start Presenting (with or without Real Data) !! 

Feed your design with data, walk through the infographic with your users from metrics to telling him about how he can use those action triggers and see your client’s reaction. The best scenario would be, ask the user to feed his real data to your chart, and before any storytelling, just present the ready chart to the user and see if he can induce the insights himself. Especially if he can identify similar action triggers as you’ve designed and even draw conclusion about it, then properly you’re working on the right track. The rest, would be up to you and your client to fine tune the whole thing and feed it back into the business chain!

Well done, Analyst! :)


So what do you think about this framework? Do you think it work for you? Or you have a better one you’ve been working on this area? Do share with us in the comment!

I hope enjoy this episode, stay tuned for part 3, and subscribe my blog or connect me if you wish to!


11:37 pm, idea-stack
[How to] Crack a Complex Business Report into Actionable Infographics (Part 2)

Three weeks after the first part, the second part is finally ready!

Last time we have talked about how to deal with the problem by starting from “something very painful” and cracking down stuff by how users read and think, this time, we put our focus into the real thing, designing and drawing.

Enjoy. :)


If you still remember the Part 1 story, my friend CB had seek for some enlightenment for dealing with a complex business dashboard with simple infographics, and with some experience in such field, I have started to crack the problem using the following workflow:

  1. Start with something *very* painful
  2. Learn no Objective but how Users Read and Think
  3. Nailing down the Metrics and Design
  4. Time to Draw Lines, Circles and Squares
  5. Design Actions Triggers with Arrows
  6. Start Presenting (with or without Real Data) !! 
(still remember our painful dashboard?)

As a short recap, we have already nailed down the scope of the data viewpoint (i.e. how the user reads) and the concept and intention behind the dashboard (i.e. how the user thinks), so it’s time for us to go further, picking the right metrics.

Step #3 Nailing down the Metrics and Design

It’s always easy for anyone to select metrics for any reporting, and if you’ve been in the management level for long, you would already know that most of the time, people simply pick ALL METRICS that they could think of. You can’t say they’re wrong, because what they’re doing is really “reporting”. Well. 

Step #3a Picking Key Metrics, Selectively

On the other hand, Analysts or Consultants would usually be more good at picking so-called “right metrics” or sometime “custom metrics”, while i believe in their expertise, still, we (i am an analyst indeed….) usually been found that using still-too-many metrics for one single problem. Think of measuring the performance of a website, how many so-called key-metrics we would have? Conversions rate? Checked. CPC? Checked. Engagement index (obviously, a custom one)? Checked. Multi-channel conversion depth? Well, sadly, checked also. Don’t blame us, because “we” really think that ALL these metrics are important (otherwise we won’t call all of them “keys”.

Trust me, such mindset is really a proficiency (and we, as consultant do proud of it too, haha), yet, we’re here to help crack a problem instead of overwhelming anyone (including ourselves), thus we shall have a different approach in our case. 

And indeed, the solution is quite simple. Again, just start from what and how users read, and pick only those existing and relevant.

(if you have forgotten how the user, my friend CB, told us how they read the dashboard, i strongly recommend you to take a visit to Part 1 again, or even read it along with Part 2 side-by-side if necessary before going further, as the story continues from where it’s ended….)
Me: So, based on how you guys read this *thing*, the Fulfillment Rate is definitely a must-have metrics?
CB: Yes, especially its trend in coming months.
Me: also the Sales?
CB: Well yes, but usually guts-feeling based on (forecast) numbers in consecutive months

Based on what the user focuses, we have already dug up two important metrics with two dimensions, product line & time (trend or in other words those “consecutive monthly numbers”):

  • Monthly YTD Fulfillment Rate (YTD-FR)
  • Monthly (Forecasted) Sales 
  • Scope: 3rd to 6th month prior to current one

Step #3b Thinking about the Graph (in advanced !)

But hey, don’t finalized anything yet (and we’re far from that indeed). If those two metrics are really that good, the original dashboard would already be doing fine! Plus, selecting the right metrics will definitely affect how we draw the graph, so don’t hesitate to ask if the user has tried anything (particularly failed attempts):

Me: and i guess you have already tried plotting this with Trend Charts?
CB: well, Dickson, no one would want to deal with a Charts with twenty-or-something trend lines……

Take notes on CB’s comments, because that’s one of the trap we always fall into: whenever we encounter something expressing in “different time period”, we have been somehow “trained” in school to use the trend chart, and in the practical world, a trend chart like this (with 20 something trends) is indeed pretty useless because its visual-complexity. Which also means that we may need another type of graph which can leverage the same set of metrics.

Now take a step backwards, what’s the true problem that CB’s facing? Reallocating resource. 
Before this? Decide allocating from which (product line) to which (another product line). 
And before this? Comparing among items before decide the first and second “which”.

Simply speaking, Comparisons between items, or a fancier word, Benchmarking.

Me: Sorry for my forgetfulness, you guys read this dashboard and make decision only once a month?
CB: Yes, and forecasts are updated *every* month.

A simple translation means: every month the trends are different, as all data except the demonstrated ones are different.

In other words, every month, the user only reads a *Snapshot* of forecasts.

Any chart is good for benchmarking as snapshot? Obvious, a Map. 

Step #3c Tuning the Metrics

If you haven’t be familiar with a Map chart already, you could spend some time on “Behind the Napkin”, as the author himself has provided a very detail explanation on how a Map works out in solving “problems in space”. Without drilling too much into that book, my personal interpretation on a Map Chart is that it provides information regarding *positions* of selected *items*, and because its geometrical nature, it is very good at visualizing the “difference” between items simply by plugging in different metrics into the two x and y axis (a.k.a. a scatter chart in excel). Such chart support very powerful framework for customization, and some popular variations, including bubble map (instead of points, we use circles with its size representing the 3rd metrics) or even the Motion Chart from Google charting api (with 4th metrics, time, included, for time-based animation) are already well used. 

As a side-story, I still remember the first bubble map chart i used (hm… actually i proposed to my previous boss who hired me as Digital Specialist without having any actual marketing experience before…. XD) is showing the different nature between social forums in Hong Kong, with volume of buzzes generated, average sentiment and engagement level as the three metrics, in order to strategize our digital marketing plan for clients. Before that the team uses “experience” or mainly “feeling” to plan for marketing efforts, yet after the map has released, we could clearly tell how to engage on which online community using what approaches based on what clients need. And after that, I got “recognized” in the first week of joining. :P

Back to our main plot, it became quite obvious for me to give Map chart a try, still we have to solve another “technical” problem, which is, unless it is a Motion Map, otherwise the Time dimension would be missing. To make sure we’re on the same track, here’s what we currently have:

  • Monthly YTD Fulfillment Rate (YTD-FR)
  • Monthly (Forecasted) Sales 
  • Scope: 3rd to 6th month prior to current one (as user can only take actual action  after two months as mentioned in Part 1)

Any method to abstract a trend into a single number? Yes, Averages. 

By using Averages (i.e. Sum of (3rd to 6th YTD-FR) / 4), we will not only be able to abstract the 3rd-6th month trend by a specific position (certain point on the map). Such calculation would also provide a similar nature as moving average (as it always takes 3rd to 6th month into account), which is always reflecting trend lagged behind the reference points (the 3rd - 6th month in this case). Lagging a forecast number is, well, sounds “usable”, so why not use it as the estimation? 

The second tuning comes to (Forecasted) Sale. Like normal retail, different product line would have different demand, pricing strategy and even marketing budget, depending on whether it is a star product and a lot other market variations. For example, Lego will always invest more in product quality for Star War series as it’s a themed line, along with the brand (Star War) popularity, usually this line sells at a premium than normal System series like City or Kingdom and of course sell more as well (hence higher revenue, or sales). To minimize such effect, particularly, to ensure we are comparing apple to apple, I have decided to change Monthly Sales into Month-to-Month (Forecasted) Sales Changes (%).

  • 4-Months Average of YTD Fulfillment Rate (3rd-6th month)
  • 4-Months Average of MoM (Forecasted) Sales Changes (3rd-6th month)

Good, so we (almost) have a Map now (with each of them representing x and y axis), but still we have one last step to go….

Step #3d Tuning the Graph

I started drafting a Map chart with those two metrics on a piece of paper along with some explanation…

(Tips: During consultation, whenever i make any “progress” in my mind, i usually ask a question “how do you feel about this?” this keeps things on the right track and of course keeps your clients think that you’re on the right track too… LOL)

Me: CB, how do you feel about this?
CB: Hm…. nice try. But how about the Fulfillment Rate? I still think that its trend is important. You remember we take notes on whether its under- or over- sales by its color ? (Red = less than 100% FR and Green = more tha or equal to 100% FR)

That was closed, I almost forgot the FIRST STEP of how the user reads the chart, the consecutive color blocks. Oh wait, color blocks? Not the numbers?

Me: Hey wait, tell me more about is the Fulfillment Rate …

By that time i know that i need to understand more regarding how the user “proceed” the information from Fulfillment Rates… so…

Me: I think a 90% FR and 70% FR are just a difference in magnitude to you (both are Red in color)?
CB: Yes.
Me: Then, what if a product has 70% 120% 70% in three consecutive month (Red, Green, Red), while the other one has 90% 90% 90% (Three Reds in-a-row), which one do you think require most of your attention?
CB: Tough one, but usually we deal with the 2nd one (with 3 Reds in-a-row) first. Why ask?

Bingo ! The missing puzzle ! By asking user’s feedback on “scenarios”, we can now tell confidently the (Fulfillment Rate) trends is somewhat more determinative to user than numbers, and it also provides us a clue and inspiration on how to bake such information along with our Map. Let’s start designing the last piece of the puzzle !

Step #3e Defining Visual Components

Remember a Map is always customizable, especially the “items” that display on it? So instead of using dots or points to indicate the position of the fulfillment-and-sales forecast of a product line on the 2D Map, how about we use something visual and indicative, hm, says, color blocks?

Each product line will have their own “Light Box” (Tips: make a name for each new thing you’ve designed, this helps to build common ground with user and ease further discussion), and each box contains 4 “Light Blocks”, and each block is either Red or Green, reflecting exactly the same color of the YTD Fulfillment Rate in 3rd to 6th (from left to right). 

Now each Box would have a total 16 combinations, and each combination will be a snapshot of the YTD-FR during 3rd to 6th month. Even better, these patterns is synchronized with users’ understanding to existing dashboard, because, indeed, we just extract the colored cells from their dashboard, right? XD

Tips: If you ask me why i pick the Light Boxes design, honestly, i really can’t tell, it somewhat like a *blink* to me… But reading more infographics from different designers would definitely help you to sharpen your intuitive in using the right shapes. A good place to begin with is of course visual.ly, where you could find a lot of cool infographics for reference. A more advanced source would be visualizing.org, some of those infographics are even interactive! Draw or sketch for whatever problems would also help strengthening and developing your own visual-solving mindset. Maybe i could write a short post regarding my practice later. :)

Me: CB, how’s that?
CB: Looks clean and effective. So what’s next?
Me: Let’s start drawing.

Step #4 Time to Draw Lines, Circles and Squares
Let’s see what we have now:
  • 4-Months Average of YTD Fulfillment Rate (3rd-6th month)
  • 4-Months Average of MoM (Forecasted) Sales Changes (3rd-6th month)
  • LightBoxes with 4-Months YTD Fulfillment Rate (3rd-6th month)
For each Product Line itme with coordinate (x,y), we can easily use the following mapping…
  • y - YTD-FR Average
  • x - MoM Sales Changes Average
  • icon - LightBox
How about the coordinates? Especially, what is the cross-ed point of two axis? Let’s think about at which point those metrics will change their “sentiment” from “negative” to “positive”:
  • y - YTD-FR Average : Range from 0% to +infinite, cells turn from Red to Green at 100%
  • x - MoM Sales Changes Average : Range from -infinite to +infinite, while a -ve changes is definitely not a good sign, and vice versa +ve changes is always good, thus 0% is the middle ground.
Obviously, a (0,1) is a good origin of the map.
Every piece of the puzzle is ready, just some more drawing…. and….
this is IT !
(I used Blue instead of Green btw.)
Finally !!
Everything is well-placed and it looks quite promising for analysis, especially now every product line has it forecasting situation nailed down on the map, and we can now easily tell which would be doing better or not  (see the arrow? no worry if you can’t figure the “why” out now, i will cover this in next part). But wait, although we’ve mapped all the items, how can we tell which product line should we reallocate resources from/to ? Is there anything more than just answering this single question from this infographics ? How can we “embed” this chart back into user’s existing business process (i.e. making decisions) ?? How to make this infographic actionable?
Next time in Part 3, we will talk about how to formulate the decisions making rules based on the infographics we have designed such that the graph itself can become part of the whole business process. We will also cover how to evolve or create alternative graphs based on user’s feedback in order to leave rooms for users to design the “final version”.
Make sure you comment below if you find yourself have learnt something here ! Subscribe me or even connect me if you wish to !!
Oh, forget to mention, it’s Chinese Lunar New Year now, so, Kung Hei Fa Choi (Good Luck & Good Fortune) !!


04:11 am, idea-stack
18 notes
[How to] Crack a Complex Business Report into Actionable Infographics (Part 1)

Three weeks ago my friend CB, who is currently working as a Logistics Analyst, has somehow given me a “quest” during a late night coffee break: to find a way to “translate” a complicated business report into infographics. Have been an Digital Analyst and practicing consultation using graphical representation for years, I had taken the chance, spent 2 hours on the problem, cracking it and visualizing it with simple charting on a piece of paper, and pens of three different colors. At the end, he looks quite satisfying, although he keeps finding ways to disprove my method with real data that he didn’t disclose to me (what a Stat. geek, still he couldn’t make it through even till now. haha.)….

Honestly i don’t know if that piece of infographic is the best solution to the problem, nor if he would put it into a good use during decision making process (i highly recommend this one though… XD), still i find that “2-hour translation” seems promising to me for future consultation, plus i somehow find no one have formulated any simply framework on this area (apart from the Napkin, yet which i think its workflow still have rooms for simplifying), so i have spent the following weeks to decode it into simple steps, and try to teach it to anyone who have an interest in cracking business problem with infographics. If you’re currently reading this, lucky you, you will be the first who learn the tricks behind (LOL). 

Let’s get started. Enjoy. :)


The workflow is indeed quite simple, yet a bit different from what you may find in the market. Traditionally people start from a “Question” from the user or client, but based on my previous experience, for most of the time the clients were either asking wrong questions (which gives us a wrong lead), or they simply don’t know what to ask, because usually the problem is too complicated for any human being to describe verbally. So how can we know what problem to tackle? Easy, start from something very painful.

Step #1 Start from something *Very* Painful

Usually, the mid-management knows this well. Remember those monthly Excel dashboards with hundreds of metrics which takes you or your junior colleagues weeks to fill all the cells and requires you, as the manager, to make important decisions based on that massive and overwhelming data? That’s exactly why my friend CB cried for help… take a look with the following spreadsheet (with mock-up data) and you will understand…. to make thing easier to understand, i use Lego’s products to replace CB’s company product lines….. (Disclosure: I’m a die-hard Lego fan !! XD)

Without a word, you should already feel how painful it is (even we know that its data is fake….).

Me: So your team need to read this every month?

CB: Yes. And we need to decide if we need to re-allocate any resources for certain product lines… (takes note on this key objective, we will talk a bit more in later section)

So a painful process followed by a even more painful decision to make, and both cannot be escaped, every month… Just like in last century (even nowadays), when people find some routine and repeating processes painful, they ask their IT to automate it. And regardless how much the automation has been done, even partially, you will start feeling yourself being more productive (and indeed you’re) because you’ve relieved yourself from a painful manually task. Infographics benefit in the same way,and particularly helpful on smoothening mental fatigue. In order to make the problem that “burying deep underground” more visible, tackling painful process would definitely a good start for us before dealing with the real problem.  

Without hesitate, i asked,

… Okay, tell me How do you read *this thing* and How do you decide an action?…

Step #2 Learn no Objectives but how Users Read and Think

Let’s start from the basic: understanding what (the hell) is going on (on the dashboard).

To make it short, here’s the break down of all essential components in the above spreadsheet that CB spent 20 minutes to explain to me…..

  • Each Row representing a product line
  • Two cells will be shown under each month (for each product), the upper with white background is Sales, while the Red/Green one is the YTD Fulfillment rate
  • Shaded areas are “Demonstrated items”, which means they’re history, actual and happened. for the rest, ALL ARE FORECASTED data (based on certain secret algorithm of their own… XD) which changed every month.
  • The Blue column is the “Current Month”
  • Sales, as named, the amount of sales achieved/forecasted of that month
  • YTD Fulfillment Rate (FR) is the key metric, it is calculated by YTD Sales over YTD Budgeted Sales, the YTD Sales would be partially “Demonstrated” and partially “Forecasted”. Budgeted Sales is somewhat a “target” that set at the last year end.
  • As manufacturing resources are planned in advanced, they’re used for the forecast of Budgeted Sales
  • the usage of YTD FR is thus obvious - how well is the current (forecasted) demand fits the original supply plan?

Me: OMG, man, you guys use *this* (the spreadsheet) every month?!

CB: Yes, we usually print it as a large poster and discuss it in the board room…

Well so now we know that It’s not just a spreadsheet, but a matrix with trends and forecasts and budget optimization flags and never-understandable prediction calculations…… how shall we kickstart? Here’s the easiest, yet most people forgot, trick: Thinking in the Users’ shoes.

Me: Well, okay. Tell me how you *read* this.

Again a summarized version from CB (well another 20 minutes…):

  1. we look for FR blocks with consecutive colors (says a chain of 4 red FR) for under/over- sales trend
  2. we then see how their Sales changes in the corresponding months
  3. finally we decide how to reallocate the (manufactory) resources from under-sold (low market demand) products to over-sold one (high market demand)
  4. we usually use this to plan for resources for the coming 6 months.

Me: How about the demonstrated items?

CB: No, they won’t affect our planning since they’re already “happened”.

Great, so not everything on the spreadsheet is needed indeed ! But don’t stop here, even though you seem already get the scope of the problem, make sure you have asked the following question at the end:

Me: So if you guys have decided to change or execute something after reading this (spreadsheet), how long it takes for the action?

CB: 2 months, that’s the minimum lead time we need to take the action.

Me: If so, can i assume the coming 1 and 2 month data may not be useful to you simply because you don’t have the time to react for them?

CB: True.

We begin getting not just a clearer picture, but a more focused scope of the problem (that actually need to be solved). And if you haven’t got confused yet (and clap for yourself if you really aren’t!), you will find that the original objective - to “re-allocate resources for certain product lines…”, is indeed too abstractive when comparing to how the user actually interacts with the dashboard (particularly, how they read, and think). This is also a typical problem that business people would frequently encounter, by cracking down a very top-level problem and then realize that it may not be solvable when down to the bottom-level. If putting Software Engineering terms into this scenario, this will be what an engineer called a proper Use Case, and building any solution around an Use Case will never go wrong because you’ve already put the problem as the focus. 

So now we’ve understood the problem, and learnt about how user interact with the dashboard, next time we will start talking about metrics selection, and of course, the design of the infographics…. ;)

Lastly, here’s a sneak peek for the coming sections ([Updated on 5th Feb] Part 2 is out now!!):

Stay tuned, and Good Night. :)



P.S. if you have any business problem, particularly have any complicated-yet-important report that you would like to “simplify”, drop me a mail at skeletonwong@gmail.com, your case may be featured on here!

*IMPORTANT NOTE: all data that me and CB have discussed, on this post and the coming ones are FAKE. My friend CB has protected his company confidentiality very well indeed, and NO REAL DATA has been leaked to anyone throughout the whole process !!*

12:59 am, idea-stack
picture HD
Background Story
Recently i have been facing some serious financial problem and thus have to find a way to introduce some cost management for myself&#8230;. and thus i have build a spreadsheet on Google Docs, storing my ATM transaction plus some special expenses items, the following is part of the by-product which looks quite fun to share here. 
                This is just a &lt;hr/&gt; line for content separation                   
How to know if yourself is overspent? Use the "EnjoyingTooMuch Index"&#160;!!ETMI =&#8221;30-day daily moving avg. on Flexible Expense&#8221; DIVIDED BY &#8220;30-day daily moving avg. on Necessity Expenses&#8221; 
(Flex Exp. includes movies, socializing, toys&#8230; stuff that without them u would somehow survive; while Nec. Exp. includes food, snack, or well for someone, internet, stuff that without them u will simply die, for good).With consecutive ETMI &lt;= 20% will be normal days, between 20% to 40% is, well, fine, and over 40% is obviously you&#8217;re &#8220;enjoying too much&#8221; (or in other words you&#8217;re working very hard to bankrupting yourself).
So how actionable is this data? Well, obviously when you that you have reached the 40% threshold for more than a week, you should definitely consider to reduce the standard of your Luxury life&#8230;. (like my graph&#8230;. Orz). Though of course, you could try to &#8220;expand&#8221; your Necessity Expenses in order to lower the index, yet it&#8217;s simply meaningless to do so because if it&#8217;s already the bottom-line of your survival, what&#8217;s the point to make yourself &#8220;more difficult to stay alive (i.e. by increasing the cost)&#8221;&#160;?&#8230; :\
So the data talks now, and obviously if my gf understand this chart and know my recent &#8220;activities&#8221; i will be already &#8220;gone for good&#8221; anyway. =_=||And yes, indeed i have already forget since when i have been an AnalyticalGeek who enjoying &#8220;translating data into graphics&#8221; (mainly solving biz problem though)&#8230;.. Orz
Anyway, so if you could think of any topics (with data) that you want me to turn into &#8220;something visual&#8221;, do drop me a line. :)
Happy Analyzing,
Dickson @ idea-stack.com

Background Story

Recently i have been facing some serious financial problem and thus have to find a way to introduce some cost management for myself…. and thus i have build a spreadsheet on Google Docs, storing my ATM transaction plus some special expenses items, the following is part of the by-product which looks quite fun to share here. 

                This is just a <hr/> line for content separation                   

How to know if yourself is overspent? Use the "EnjoyingTooMuch Index" !!

ETMI =”30-day daily moving avg. on Flexible Expense” DIVIDED BY “30-day daily moving avg. on Necessity Expenses” 

(Flex Exp. includes movies, socializing, toys… stuff that without them u would somehow survive; while Nec. Exp. includes food, snack, or well for someone, internet, stuff that without them u will simply die, for good).

With consecutive ETMI <= 20% will be normal days, between 20% to 40% is, well, fine, and over 40% is obviously you’re “enjoying too much” (or in other words you’re working very hard to bankrupting yourself).

So how actionable is this data? Well, obviously when you that you have reached the 40% threshold for more than a week, you should definitely consider to reduce the standard of your Luxury life…. (like my graph…. Orz). Though of course, you could try to “expand” your Necessity Expenses in order to lower the index, yet it’s simply meaningless to do so because if it’s already the bottom-line of your survival, what’s the point to make yourself “more difficult to stay alive (i.e. by increasing the cost)” ?… :\

So the data talks now, and obviously if my gf understand this chart and know my recent “activities” i will be already “gone for good” anyway. =_=||

And yes, indeed i have already forget since when i have been an Analytical
Geek who enjoying “translating data into graphics” (mainly solving biz problem though)….. Orz

Anyway, so if you could think of any topics (with data) that you want me to turn into “something visual”, do drop me a line. :)

Happy Analyzing,

Dickson @ idea-stack.com

12:02 pm, idea-stack
picture HD
That&#8217;s how PR People handle Crisis nowadays.
Lesson learnt.

That’s how PR People handle Crisis nowadays.

Lesson learnt.