Context. The onset of massive star formation is not well understood because of observational and theoretical difficulties. To find the dense and cold clumps where massive star formation can take place, we compiled a sample of high infrared extinction clouds. We observed the clumps in these high extinction clouds in the 1.2 mm continuum emission and ammonia with the goals of deriving the masses, densities, temperatures, and kinematic distances. Aims. We try to understand the star-formation stages of the high extinction clumps by studying their infall and outflow properties, the presence of a young stellar object (YSO), and the level of the CO depletion. Are the physical parameters, density, mass, temperature, and column density correlated with the star-forming properties? Does the cloud morphology, quantified through the column density contrast between the clump and the clouds, have an impact on the evolution of star formation occurring inside it? Methods. Star-formation properties, such as infall, outflow, CO depletion, and the presence of YSOs, were derived from a molecular line survey performed with the IRAM 30 m and the APEX 12 m telescopes. Results. We find that the HCO + (1-0) transition is the most sensitive for detecting infalling motions. SiO, an outflow tracer, was mostly detected toward sources with infall, indicating that infall is accompanied by collimated outflows. We calculated infall velocities from the line profiles and found them to be of the order of 0.3-7 km s −1 . The presence of YSOs within a clump depends mostly on the clump column density; no indication of YSOs were found below 4 × 10 22 cm −2 . Conclusions. Star formation is on the verge of beginning in clouds that have a low column density contrast; the infall is not yet present in the majority of the clumps. The first signs of ongoing star formation are broadly observed in clouds where the column density contrast between the clump and the cloud is higher than two; most clumps show infall and outflow. Finally, we find the most evolved clumps in clouds that have a column density contrast higher than three; in many clumps, the infall has already halted, and toward most clumps we found indications of YSOs. Hence, the cloud morphology, based on the column density contrast between the cloud and the clumps, seems to have a direct connection with the evolutionary stage of the objects forming inside.