How can you reactivate your training video library?

1. Identify what is getting in the way of learning

A video with a low com­ple­tion rate is not ne­ces­sar­ily a poor video. Learners may be look­ing for a spe­cif­ic answer, be short of time or already know part of the sub­ject. Con­versely, fin­ish­ing a video does not prove that someone can apply its con­tent. Ex­am­ine feed­back and the task to be per­formed before de­cid­ing to remake everything.

• Is the con­tent still ac­cur­ate, up to date and ap­pro­pri­ate for the audi­ence?
• What work­place task should learners be able to per­form after the module?
• At which points do ques­tions, errors or dro­pouts occur?
• Is there an activ­ity that checks un­der­stand­ing beyond simply watch­ing?

2. Connect videos to a concrete objective

Set an ob­serv­able ob­ject­ive: “identi­fy the doc­u­ments to re­quest during an ini­tial meet­ing”, rather than “un­der­stand cus­tom­er re­la­tion­ships”. Then clas­si­fy the videos: keep, cor­rect or remove. Identi­fy the pas­sages es­sen­tial to that ob­ject­ive and the op­tion­al ma­ter­i­al for fur­ther ex­plor­a­tion.

Out­dated ma­ter­i­al must be cor­rec­ted before an AI uses it as a ref­er­ence. Adding a con­ver­sa­tion to a video does not fix an in­cor­rect in­struc­tion or an ob­sol­ete pro­ced­ure. Ap­point someone to ap­prove the con­tent and record the ver­sion used.

1. Write down the ex­pec­ted action and its suc­cess cri­ter­ia.
2. Con­nect each se­quence to a concept or a step in that action.
3. Pre­pare an ex­er­cise that makes prac­tic­al ap­plic­a­tion ob­serv­able.

3. Organise sequences without rerecording everything

Divide the con­tent or mark changes of topic with ex­pli­cit head­ings. A chapter should help learners find an answer, not simply match an ar­bit­rary dur­a­tion. A com­plex demon­stra­tion may take longer than a defin­i­tion: there is no uni­ver­sal length suit­able for every course. To con­sider the role of di­git­al learn­ing in your of­fer­ing, also read our art­icle on training providers and digital learning paths.

Seg­ment­ing con­tent and high­light­ing im­port­ant in­form­a­tion are among the prin­ciples Cyn­thia Brame de­scribes in her review of edu­ca­tion­al videos [1]. These re­com­mend­a­tions sup­port learn­ing design; they do not guar­an­tee a quan­ti­fied im­prove­ment for your lib­rary.

4. Combine spoken questions with practice

Let learners ask for an explanation when they need it

With Di­dacticBot, learners can ask a ques­tion aloud during train­ing. The video pauses; the AI draws on ap­proved learn­ing ma­ter­i­als, and the learn­ing path re­sumes when re­ques­ted. This in­ter­ac­tion lets learners ask for an ex­plan­a­tion or re­ph­ras­ing without leav­ing the con­text of the course.

The conversational demonstration on Qualiopi, France’s train­ing qual­ity cer­ti­fic­a­tion frame­work, shows how this works for a spe­cif­ic topic. For your lib­rary, im­port­ing, struc­tur­ing and track­ing ar­range­ments must be con­firmed in a demon­stra­tion tailored to your needs. Ex­plore the LMS and voice AI features to dis­tin­guish the ex­amples shown from cap­ab­il­it­ies that need to be con­firmed.

Ask learners to retrieve information, then use it

After a se­quence, first ask learners to for­mu­late an answer without watch­ing the video again, then provide ex­plan­at­ory feed­back. Next, offer a situ­ation in which they must make a de­cision or take action. The review by Dun­losky and col­leagues iden­ti­fies re­triev­al prac­tice and dis­trib­uted prac­tice as useful tech­niques in the con­texts stud­ied [2]. It does not meas­ure Di­dacticBot’s ef­fect­ive­ness.

A quiz checks cer­tain learn­ing out­comes; a scen­ario-based activ­ity ex­am­ines how they are used. DidacticBot serious games can il­lus­trate this second ap­proach. Choose a game based on the learn­ing ob­ject­ive, rather than re­pla­cing a rel­ev­ant activ­ity with a game purely for nov­elty.

5. Before-and-after example: preparing for a client meeting

This fic­tion­al scen­ario il­lus­trates the method: a train­ing pro­vider has a video present­ing the stages of an ini­tial client meet­ing. This ex­ample makes no claims about client res­ults.

6. Evaluate the pilot before rolling it out

Keep a baseline from before the changes: the same ob­ject­ives, a com­par­able audi­ence and an ex­er­cise as­sessed against the same cri­ter­ia. After the trial, dis­tin­guish ex­per­i­ence feed­back, view­ing activ­ity and learn­ing out­comes. If groups or con­di­tions differ, do not present their com­par­is­on as causal evid­ence.

• Un­der­stand­ing: which errors occur in the ex­er­cise, and are they cor­rec­ted?
• Use: can learners find the right se­quence and resume the learn­ing path?
• Ques­tions: which points need a better ex­plan­a­tion or in­ter­ven­tion from the train­er?
• Main­ten­ance: who cor­rects the ma­ter­i­als and re­approves the module when the pro­ced­ure changes?

Set suc­cess cri­ter­ia before the trial. The pro­ject’s dur­a­tion de­pends on the con­di­tion of the videos, the ad­apt­a­tions needed and the avail­ab­il­ity of re­view­ers. No uni­ver­sal dead­line, target com­ple­tion rate or return on in­vest­ment is prom­ised here. Training providers can use this grid to scope their pro­ject.

Decide whether to repair, supplement or replace

Before fund­ing new pro­duc­tion, dis­tin­guish a prob­lem with the con­tent from a prob­lem ac­cess­ing it. An ac­cur­ate ex­plan­a­tion that is hard to find is more likely to need better sign­post­ing. A demon­stra­tion show­ing an in­ter­face that has changed may need a tar­geted sup­ple­ment. A pro­ced­ure that is now in­cor­rect must be re­moved from the active learn­ing path, even if film­ing it was ex­pens­ive. Money already spent does not jus­ti­fy re­tain­ing a mis­lead­ing re­source.

Also con­sider how stable the sub­ject is: an en­dur­ing prin­ciple and a fre­quently chan­ging soft­ware screen do not have the same main­ten­ance needs. For the latter, sep­ar­at­ing the ex­plan­a­tion of the prin­ciple from the op­er­at­ing in­struc­tions makes later cor­rec­tions easier. This ed­it­or­i­al de­cision does not re­quire any par­tic­u­lar player fea­ture; it can be im­ple­men­ted through two sep­ar­ate re­sources and ex­pli­cit links.

The ex­pec­ted de­liv­er­able is a de­cision sheet for each re­source: the prob­lem iden­ti­fied, its impact on the task, the chosen solu­tion, the person re­spons­ible for ap­prov­al and the con­di­tion for with­draw­al. Add the usage rights that need check­ing, par­tic­u­larly when presenters, third-party images or client doc­u­ments appear. An in­ab­il­ity to reuse ma­ter­i­al may change the de­cision, re­gard­less of its edu­ca­tion­al qual­ity.

Address what video alone does not make accessible

Bring­ing videos back into use is also an op­por­tun­ity to check in­form­a­tion that is only shown or only spoken. The W3C WAI guide to accessible media dis­tin­guishes cap­tions, tran­scripts, de­scrip­tions of visual in­form­a­tion and player cap­ab­il­it­ies. These com­pon­ents ad­dress dif­fer­ent needs; simply adding text below a video does not demon­strate that the entire learn­ing path is ac­cess­ible.

In a demon­stra­tion, identi­fy ac­tions re­ferred to as “here”, col­ours used as the only cue and values dis­played without being read aloud. Pre­pare a de­scrip­tion that makes the action un­der­stand­able without seeing the point­er pre­cisely. Review tech­nic­al cap­tions: a miss­ing neg­at­ive or an in­cor­rectly tran­scribed spe­cial­ist term can re­verse an in­struc­tion. Also check that cor­rec­tions stay aligned with video ver­sions.

The ac­cept­ance-test­ing record should con­nect each ob­served bar­ri­er to a cor­rec­tion and an actual check: access to the al­tern­at­ive ma­ter­i­al, un­der­stand­ing the in­struc­tions and nav­ig­a­tion through the learn­ing path. Con­firm what the chosen player can do rather than as­sum­ing its cap­ab­il­it­ies. A sup­ple­ment­ary doc­u­ment may help as a stop­gap, but it does not re­place an ana­lys­is of the needs of the people con­cerned.

Manage the transition without leaving two conflicting versions in circulation

In this case, ac­cept­ance test­ing means more than watch­ing the new video. The re­view­er searches for the old form through the usual entry points: the lib­rary, doc­u­ment area and launch mes­sages. They check that links lead to the se­lec­ted ver­sion and that the sup­ple­ment­ary sheet car­ries the same pro­ced­ure ref­er­ence. If an old re­source must remain archived, its status must be clear and dis­tinct from the active learn­ing path.

Fi­nally, pre­pare a re­lease check­list: re­sources pub­lished, re­sources with­drawn, links checked, out­stand­ing issues and the person au­thor­ised to sign off those issues. Dis­tin­guish a block­ing issue, such as an in­cor­rect in­struc­tion, from a present­a­tion im­prove­ment that can wait. Define the next review trig­ger, such as a change to the form, rather than making a vague prom­ise of reg­u­lar up­dates. This check­list com­ple­ments edu­ca­tion­al eval­u­ation: it safe­guards doc­u­ment con­sist­ency without claim­ing to meas­ure learn­ing out­comes.

Tell train­ers about the changes so they do not con­tin­ue re­com­mend­ing an old link. Keep a record of this com­mu­nic­a­tion in the re­lease doc­u­ment­a­tion.

Key takeaway: reactivating does not mean remaking everything

Keep re­li­able con­tent, con­nect it to a con­crete action, and add op­por­tun­it­ies to ask ques­tions and prac­tise. Start with one module, ob­serve what learners can ac­tu­ally do suc­cess­fully, and im­prove the learn­ing path before ex­tend­ing it. Voice con­ver­sa­tion is one way to sup­port un­der­stand­ing; it com­ple­ments in­struc­tion­al design and the train­er’s role.