Sign in




1. What is Speech Intelligence Analytics™?

A proprietary speech engine, designed to measure, and rate the quality of selected speech events.

Society gives many labels to describe the act of human communication ( 'keynote', 'presentation', 'interview', 'meeting', 'discussion panel'). gweek looks (and listens) through these labels and scrutinises the denominators common to them all. After all, a listener uses the same mind with which to feel attracted to, or distracted from, a speaker. No matter the scenario.

We believe that a ratings system around a person's intrinsic communication competency will lead to more informed, transparent decision-making. One score helps you see that person beyond the emotional halo. Beyond their current reputation. And if you don't yet have a reputation, your own communication rating can help establish yours on a level playing field.

2. Is '100' the best rating?

Not to confuse with the gweek100, which is our global ranking platform, a gweek score is out of '100'. You can always check the global and sector averages at gweekdash. 

The closer you are to '100', the better your balance of substance with authenticity. gweek is also about collecting big data in speech; the more we gweek, the more we understand about human speech productivity. 

3. Welcome to "g"- at the heart of SIA™

Your gweek “g” comprises embedded layers of communication appraisal. It’s our definition of leadership communication.

1. Consistency “c” - no matter the setting, we use the same criteria to analyse the speaker and look for signs of the ‘real you’ coming across consistently, despite the variety of settings your profession may demand. “c” is tracked through deriving your average speech score, over time.

2. Productivity “p” (vs. co-ordination “c-o”)- signs of impactful communication, both emotional as well as intellectual, pull a speech score up; specific lack in verbal and non-verbal co-ordination pull it down. “p” is a physical way of describing the degree to which your style, or personality, actually comes across. We can demonstrate a correlation between “p” and “f”; on average, that a higher “c-o” has an exponentially higher “p”.

3. Substance and authenticity “s” and “a” are subsets of “p”- listeners can realistically make decisions on you through how you communicate with them. The more “c” and “p”, the easier this becomes for the listener. Are you too raw? Resulting in higher “a”, but lower “s”. Or are you over-trained or scripted? Resulting in higher “s” but lower “a”. We look for a balance. More than anything, this reflects your skill in communicating with those looking to you for leadership. The real you. We do not directly measure choice in word or syntactical arrangement. But we do indirectly track both, and we directly track everything else. “g” = “c” = “p” over time (where “p” is defined as “s” and “a” in balance).

Raw data is tracked and collected by a trained speech technician, as a speaker is 'live' in action, or through high quality audio-video of the speaker.

Our proprietary speech engine, Speech Intelligence Analytics™ (SIA™) interprets the collected data streams. SIA™ derives speech scores based on selected SIA™ level.

Track the ups and downs of your chosen personality through their speech career. Aspire to get your own gweek and rank in the gweek100.

4. gweekology explained

  • Authenticity: the listener's need to sense that the look and the sound of the speaker is for real. Amongst others, we observe the following instances where authenticity is diminished: 1) the speaker is clearly scripted, speaking but not actually recalling memories 2) the speaker's word-by-rate speech rate is sustained at a lower speed than normal (usual when speaker is scripted) 3) the speaker maintains unnaturally prolonged eye contact with the listener/s (see Cameron 'debate'), or to the contrary does not make any eye contact (see Zuma 'keynote') 4) the speaker's comments are all 'single' ideas (see 'substance')
  • Comment: 1) a single uninterupted utterance, containing both words and sounds. 2) Comments are identifiable by periods of silence or empty pauses occuring before and after. For example, the following is a 6-unit comment: "Oh you're wearing your new shoes." 3) Comments can contain single or compound ideas. Or indeed, none. The above example is deemed a single idea comment. 4) Speech comments are voiced in a context-rich environment and behave differently to written language sentences, which require more 'units' to set the scene and actual meaning. In short, a comment can enjoy fewer 'units' than a written sentence. Using both verbal and non-verbal behaviours, comments can draw the listener's attention when specific emphasis sits over a particular 'word'. Hence the listener can focus on the sound of two units "new shoes" in the above example, to understand whether the speaker is expressing surprise, annoyance or delight at the wearing of the new shoes.
  • Co-ordination, in speech: 1) Defined by the absence of disfluencies or delays in any given speech comment. A disfluency is characterised by any or a multiple of the following: filled pause, repair, repeat, word-to-word hesitation 2) Speech co-ordination has different characteristics to written language fluency. gweek does not scan for syntactical or 'grammatical' completion. Hence "I woke up this morning with a dreadful..." is deemed a co-ordinated comment. That the speaker does not complete the explicit meaning with "headache" is beside the point. The listener, by reading the voice tone of the comment so far has probably already guessed the 'key word' and doesn't need to hear it. They deduce the meaning ahead of the speaker completing it. This is called 'top down' analysis. The equivalent written sentence would require an explicit completion of the meaning and is less flexible in this respect. See 'comment'.
  • get your gweek: professional speech analysis via SIA™; 3 speech scores awarded by gweek are then averaged to become your speech rating, your gweek. A range of scenarios will better reflect your broader capabilities e.g. keynote, discussion, debate and media interview. A competitive gweek could qualify you for gweek100;
  • gweek (noun): 1) A current and personal speech rating; calculated as the average of the speech scores recorded to date. e.g. "My gweek's 82. What's yours?" 2) A minimum of 3 speech scores is suggested to calculate your gweek, and from a diversity of speech events 3) Only gweek is qualified to issue a gweek.
  • gweek (verb): To observe and record the speech behaviours of a speaker 2) A gweek speech technician is trained and continually assessed in the science of speech/communication data observation and recording.
  • Speech score: the selected SIA™ algorithm will calculate the score for the speech episode/event in question. Most SIA™ levels require 60 seconds of continuous data. Others 120 seconds.
  • Speech technician: an individual trained to identify and record speech data.
  • Substance: Speaker's ability to generate meaningful ideas; gweek defines as either a simple or a compound comment. A 'simple' will contain up to one idea (e.g. "I'm the right person for the job"); a 'compound' has up to two (e.g. "I'm the right person for the job and I'm keen to show you why."). We deem the balanced use of compounds as verbally more impressive than a whole string of singles."). The concept of 'simples' and 'compounds' arises over both 1) the speaker's ability to remain productive and 2) the listener's 'at the time' capacity to store the comments in short-term memory. Hence, as speakers we might stay away from comments that contain no meaning (either intellectual or emotional) and comments that contain more than 2 ideas within (too complex for the listener to store and then act on.)
  • Productivity, in speech: The degree to which a comment is likely to provoke substantive thought or feeling in the listener, within the context of the communication. gweek considers an array of both verbal and non-verbal behaviours in determining the extent of productivity, from comment composition/substance (verbal) to eye movement (non-verbal).
  • Voice pitch: or modulation, or frequency or emphasis. As with any phenonemon of delivery, we are describing the features that occur within the boundaries of a single comment. We observe that speakers put emphasis over some words (and not others); to help listeners extract important meanings from any comment. Naturally, we seem to keep comments to the right length, to help us select the right emphasis. Thus in the comment "Raining?" The pitch of this single word comment helps us understanding whether the speaker is genuinely surprised, or genuinely frustrated. We use pitch to adjust the emotional meaning. "OH NO I'm late for the BUS".
  • Voice pitch set and reset: A complex phenonemon related to voice pitch, or modulation, or fequency or emphasis. As with any phenonemon of delivery, we are describing the features that occur within the boundaries of a single comment. And whether the verbal (words) are co-ordinated with the non-verbal (sound and look). Pitch set describes how we organise our personal range of sounds so that the listener knows how we feel (in any given comment). To emphasise some words and de-emphasive others. Like a musical scale, speech uses a fixed range of sound signals to help narrow the listener's interpretation of our emotional emphasis and meaning . To recycle and re-use the same range, our pitch has to 'reset' during the silence between spoken comments. Without 'reset' there can be no meaningful range of sounds to help a listener interpret both the meaning of the comment and how the speaker feels. 

5. Which SIA™ level does the gweek100 global ranking employ?

SIA3™ - speech delivery (co-ordination and productivity)

6. Why SIA3™?

The components of SIA3™ allow for analysis of both substance and authenticity