Question

Topic: E-Marketing

Generate Reports

Posted by Anonymous on3/23/2006 at 2:04 PM ET 250 Points
I've just joined as Market Analyst in a e-marketing Co. Responsible for generating reports, need help and suggestions
To continue reading this question and the solution, sign up ... it's free!
SUBSCRIBE
Or,SIGN IN

RESPONSES

  • Posted byROIHUNTERon3/26/2006 at 8:49 AM Accepted
    lilly.josiah

    Your question is very broad, so I am left with providing a very broad answer. I will answer this from the point of view of a new hire.

    1. Understand all the feedback systems of your company or clients. This will help you determine what type of campaigns you can run with feedback systems already in place.
    2. Understand what type of campaigns you will be responsible for creating or monitoring in the next 6 months.
    3. Determine if any new feedback systems need to be created before the new campaigns begin.

    If you would like to ask a more specific question, maybe we could provide more insight.

    Hope that helps,
  • Posted on4/7/2006 at 12:30 PM Author
    I agree my question is braod, but still the answers received are very apt and relevant, Thanks a bunch to all who have answered
  • Posted bysteven.alkeron4/8/2006 at 7:28 PM Accepted
    Dear lilly.josia

    Part 1 – The Background

    Firstly, welcome to MarketingProfs – I hope that we can give you some insights into your question. The problem with breadth in a question tends to be length in the answer and the risk that 90% of that written will be of little relevance. Hence the reason why this is in two parts.

    我是一个CRM从业者(还有其他一些东西)one of the key parts to making a system pay for itself is to encourage the users to record what happens and then to report on it. Both aspects are taxing in their own right and you can’t do the latter without the former. Interpreting reports and taking actions is the part of art of management, but that’s another issue.

    Recording can appear arduous, but with e-marketing, the basic data can often be tracked automatically and logged by the systems you have in place. If that is not currently the case with your company, the sooner you change your practices the better. Recording is essential because it is the means to measurement and if you can’t measure something, you can’t report on it, manage it or improve on it.

    What do you need to measure is the next question which needs an answer. Because you haven’t stated any specifics, I can’t advise you on this in a few hundred words, so if in doubt, list the data which you are able to record and measure and log the lot. You can safely ignore data you have recorded which has no relevance to the reports you require, but you cannot produce a report or analyse statistics if the base data is unavailable.

    Let’s try to get one basic correct. Measurement and recording is the process of producing data. Data is extremely useful in the way that a list of phone numbers can be useful but until someone connects them to some names and addresses, i.e. a phone book.

    And that phone number data does not become useful information until someone does something with it, such as making sales calls to a set of addresses, or looking up Aunty Minnie’s phone number. If you want something which is of value from a marketing information standpoint, derived from the base data (Phone Numbers) you need to cross correlate them with the addresses against which you can then cross reference some socio-economic data by matching the post codes. This is a list on a database and can be the subject of a paper wasting report.

    You can then move on to making some sales use of the data by selecting those phone numbers in an AB12 sector, running a report on it and then attempting to sell them something expensive by bugging the wits out of them with cold calls.

    Reporting is used to produce the list of appropriate prospects, but you don’t get onto management reporting until you analyse how effective your sales efforts are, by a number of different criteria, as a result of all those phone calls.

    Steve Alker
    Unimax Solutions
  • Posted bysteven.alkeron4/8/2006 at 7:29 PM 米ember
    Dear lilly.josia (Again)

    Part 2 – As Applied to your company

    The chain you are looking at in your company is probably very similar. You will record things in order to measure them. You will then work out how to analyse them and report on them in order extract some information. From the information, you will then set further tasks, all of which will be measurable in order to extract some value and to see how effective you are at adding value.

    You will need to report on that last set of activities in order to manage an outcome.

    So, record, measure, tabulate, manipulate, analyse, report, do something pro-active, report again, analyse again and then manage your enterprise and people according to the results of the analysis which derives from the reports.

    Too many companies try to manage from the data, which is about as useful as trying to save a patients life by looking at 60 pages of biometric figures from an intensive care ward monitor. The figures need to be categorised, manipulated, analysed and reported on in order for a doctor to proceed with an appropriate therapy. Then the therapy has to be monitored (More data, more information) and the outcome judged, the treatment adjusted and the prognosis forecasted.

    Sorry about all the analogies but the lack of information in the question almost mandates example ay parable. Once you are recording everything that you can record to produce the data, you then need to decide which metrics you want to look at.

    These are usually gross and net numbers, by volume and by $ or £ value. They will include costs incurred and therefore margin. You will accumulate all of the ongoing data relating to the parameters kwhite mentions and then some more. You will look up market data and other external factors which could affect your performance and record these as well,

    Next you will want to manipulate your data with some statistical techniques. These might be simplistic like regressions and calculating confidence factors or, depending on the quantity of data you might go for something more adventurous such as grouping and chi squared analysis. For grouping, you could look to break down your data by association of summary information against identifiable categories. Two dimensional pivot tables are useful here, but with all the variables in eBusiness and factoring in external market information you will want to study n-dimensional pivot data in order to glean some information you can report on. Tableau software is pretty good at doing this for sales and operational data whereas Excel will take you weeks.

    You will need to look at a lot of ratios. You will also be asked to compare these ratios with other’s you can glean from within your industry. This is total rubbish as the only company to which your figures, statistics and ratio’s pertain is your own – there’s piles of info on the fallacy of “comparison with the norm” and other such tripe in other postings here on MarketingProfs.)

    接下来你将不是看一些预测predictive work and report on these. Linear extrapolation is the usual tool and can be achieved from the earlier analysis, but if your data model is complex, you will need to look to more advanced techniques. Here, you may well find yourself stymied in predicting trends or forecasts because simple regression and extrapolation fails due to the apparent variability of the data.

    If you don’t know the relationship between basic sales and marketing input data and the desired results (Sales, page impressions, reply emails, web forms, click-throughs etc) then you could utilise a Fuzzy Logic or Neural product to find the associations, groupings and eventually the relationship between your base data and the forecast or predictive results you are seeking.

    I’m just looking at NeuralWare and its suit Predict, where for the last few weeks, I’ve been feeding in data from web-sites, email campaigns and other marketing figures and it appears to build some very solid associations and data handling models to predict what is likely to be a forecast or a pattern of future behaviour.

    I was sceptical at first about having a piece of software working out which set of algorithms to use to handle my data (Rather than having me guess at one) but it works very, very well. At the end of the process I can show that it will accurately predict the actual sales figures as a forecast based on historic data. The last 2 years of enquiry, quote and order stats successfully predicted the figure we actually achieved 6 months ago – ergo I have some confidence that the last 3 years of figures to date, will give me a reasonable forecast for the next 1, 3, 6 and 12 months. I don’t know what the hell the equations are which it has developed and used, but I do know that has tested them and inferred results which were bourn out of historic data could not give the to the model because they havn’t happened yet!

    One last point – this has all been about metrics, analysis, forecasts and management fiscal information, current, past and predicted. Don’t forget to put in your reports some of your feelings and some subjective views taken from leaders in your company and in your industry (Meet them at a trade show and ask them) I’ve said it before but it’s worth repeating, the business world does not solely depend on hard wired Cartesian geometry, statistical analysis and mathematical reporting. Your human insights are of great value as well, so record them and be prepared to justify them.

    Steve Alker
    Unimax Solutions
BACK TO TOP

Post a Comment

相关的讨论

Community Info

Top 25 Experts
(E-Marketing)